Annotation of gforth/gforth.ds, revision 1.18

1.1       anton       1: \input texinfo   @c -*-texinfo-*-
                      2: @comment The source is gforth.ds, from which gforth.texi is generated
                      3: @comment %**start of header (This is for running Texinfo on a region.)
1.4       anton       4: @setfilename gforth.info
1.17      anton       5: @settitle Gforth Manual
1.4       anton       6: @comment @setchapternewpage odd
1.1       anton       7: @comment %**end of header (This is for running Texinfo on a region.)
                      8: 
                      9: @ifinfo
1.17      anton      10: This file documents Gforth 0.1
1.1       anton      11: 
1.17      anton      12: Copyright @copyright{} 1994 Gforth Development Group
1.1       anton      13: 
                     14:      Permission is granted to make and distribute verbatim copies of
                     15:      this manual provided the copyright notice and this permission notice
                     16:      are preserved on all copies.
                     17:      
1.4       anton      18: @ignore
1.1       anton      19:      Permission is granted to process this file through TeX and print the
                     20:      results, provided the printed document carries a copying permission
                     21:      notice identical to this one except for the removal of this paragraph
                     22:      (this paragraph not being relevant to the printed manual).
                     23:      
1.4       anton      24: @end ignore
1.1       anton      25:      Permission is granted to copy and distribute modified versions of this
                     26:      manual under the conditions for verbatim copying, provided also that the
                     27:      sections entitled "Distribution" and "General Public License" are
                     28:      included exactly as in the original, and provided that the entire
                     29:      resulting derived work is distributed under the terms of a permission
                     30:      notice identical to this one.
                     31:      
                     32:      Permission is granted to copy and distribute translations of this manual
                     33:      into another language, under the above conditions for modified versions,
                     34:      except that the sections entitled "Distribution" and "General Public
                     35:      License" may be included in a translation approved by the author instead
                     36:      of in the original English.
                     37: @end ifinfo
                     38: 
                     39: @titlepage
                     40: @sp 10
1.17      anton      41: @center @titlefont{Gforth Manual}
1.1       anton      42: @sp 2
1.17      anton      43: @center for version 0.1
1.1       anton      44: @sp 2
                     45: @center Anton Ertl
1.17      anton      46: @sp 3
                     47: @center This manual is under construction
1.1       anton      48: 
                     49: @comment  The following two commands start the copyright page.
                     50: @page
                     51: @vskip 0pt plus 1filll
1.17      anton      52: Copyright @copyright{} 1994 Gforth Development Group
1.1       anton      53: 
                     54: @comment !! Published by ... or You can get a copy of this manual ...
                     55: 
                     56:      Permission is granted to make and distribute verbatim copies of
                     57:      this manual provided the copyright notice and this permission notice
                     58:      are preserved on all copies.
                     59:      
                     60:      Permission is granted to copy and distribute modified versions of this
                     61:      manual under the conditions for verbatim copying, provided also that the
                     62:      sections entitled "Distribution" and "General Public License" are
                     63:      included exactly as in the original, and provided that the entire
                     64:      resulting derived work is distributed under the terms of a permission
                     65:      notice identical to this one.
                     66:      
                     67:      Permission is granted to copy and distribute translations of this manual
                     68:      into another language, under the above conditions for modified versions,
                     69:      except that the sections entitled "Distribution" and "General Public
                     70:      License" may be included in a translation approved by the author instead
                     71:      of in the original English.
                     72: @end titlepage
                     73: 
                     74: 
                     75: @node Top, License, (dir), (dir)
                     76: @ifinfo
1.17      anton      77: Gforth is a free implementation of ANS Forth available on many
1.1       anton      78: personal machines. This manual corresponds to version 0.0.
                     79: @end ifinfo
                     80: 
                     81: @menu
1.4       anton      82: * License::                     
1.17      anton      83: * Goals::                       About the Gforth Project
1.4       anton      84: * Other Books::                 Things you might want to read
1.17      anton      85: * Invocation::                  Starting Gforth
                     86: * Words::                       Forth words available in Gforth
1.4       anton      87: * ANS conformance::             Implementation-defined options etc.
1.17      anton      88: * Model::                       The abstract machine of Gforth
                     89: * Emacs and Gforth::            The Gforth Mode
1.4       anton      90: * Internals::                   Implementation details
                     91: * Bugs::                        How to report them
1.17      anton      92: * Pedigree::                    Ancestors of Gforth
1.4       anton      93: * Word Index::                  An item for each Forth word
                     94: * Node Index::                  An item for each node
1.1       anton      95: @end menu
                     96: 
                     97: @node License, Goals, Top, Top
                     98: @unnumbered License
                     99: !! Insert GPL here
                    100: 
                    101: @iftex
                    102: @unnumbered Preface
1.17      anton     103: This manual documents Gforth. The reader is expected to know
1.1       anton     104: Forth. This manual is primarily a reference manual. @xref{Other Books}
                    105: for introductory material.
                    106: @end iftex
                    107: 
                    108: @node    Goals, Other Books, License, Top
                    109: @comment node-name,     next,           previous, up
1.17      anton     110: @chapter Goals of Gforth
1.1       anton     111: @cindex Goals
1.17      anton     112: The goal of the Gforth Project is to develop a standard model for
1.1       anton     113: ANSI Forth. This can be split into several subgoals:
                    114: 
                    115: @itemize @bullet
                    116: @item
1.17      anton     117: Gforth should conform to the ANSI Forth standard.
1.1       anton     118: @item
                    119: It should be a model, i.e. it should define all the
                    120: implementation-dependent things.
                    121: @item
                    122: It should become standard, i.e. widely accepted and used. This goal
                    123: is the most difficult one.
                    124: @end itemize
                    125: 
1.17      anton     126: To achieve these goals Gforth should be
1.1       anton     127: @itemize @bullet
                    128: @item
                    129: Similar to previous models (fig-Forth, F83)
                    130: @item
                    131: Powerful. It should provide for all the things that are considered
                    132: necessary today and even some that are not yet considered necessary.
                    133: @item
                    134: Efficient. It should not get the reputation of being exceptionally
                    135: slow.
                    136: @item
                    137: Free.
                    138: @item
                    139: Available on many machines/easy to port.
                    140: @end itemize
                    141: 
1.17      anton     142: Have we achieved these goals? Gforth conforms to the ANS Forth
                    143: standard. It may be considered a model, but we have not yet documented
1.1       anton     144: which parts of the model are stable and which parts we are likely to
1.17      anton     145: change. It certainly has not yet become a de facto standard. It has some
                    146: similarities and some differences to previous models. It has some
                    147: powerful features, but not yet everything that we envisioned. We
                    148: certainly have achieved our execution speed goals (@pxref{Performance}).
                    149: It is free and available on many machines.
1.1       anton     150: 
                    151: @node Other Books, Invocation, Goals, Top
                    152: @chapter Other books on ANS Forth
                    153: 
                    154: As the standard is relatively new, there are not many books out yet. It
1.17      anton     155: is not recommended to learn Forth by using Gforth and a book that is
1.1       anton     156: not written for ANS Forth, as you will not know your mistakes from the
                    157: deviations of the book.
                    158: 
                    159: There is, of course, the standard, the definite reference if you want to
                    160: write ANS Forth programs. It will be available in printed form from
                    161: Global Engineering Documents !! somtime in spring or summer 1994. If you
                    162: are lucky, you can still get dpANS6 (the draft that was approved as
                    163: standard) by aftp from ftp.uu.net:/vendor/minerva/x3j14.
                    164: 
                    165: @cite{Forth: The new model} by Jack Woehr (!! Publisher) is an
                    166: introductory book based on a draft version of the standard. It does not
                    167: cover the whole standard. It also contains interesting background
                    168: information (Jack Woehr was in the ANS Forth Technical Committe). It is
                    169: not appropriate for complete newbies, but programmers experienced in
                    170: other languages should find it ok.
                    171: 
                    172: @node Invocation, Words, Other Books, Top
                    173: @chapter Invocation
                    174: 
                    175: You will usually just say @code{gforth}. In many other cases the default
1.17      anton     176: Gforth image will be invoked like this:
1.1       anton     177: 
                    178: @example
                    179: gforth [files] [-e forth-code]
                    180: @end example
                    181: 
                    182: executing the contents of the files and the Forth code in the order they
                    183: are given.
                    184: 
                    185: In general, the command line looks like this:
                    186: 
                    187: @example
                    188: gforth [initialization options] [image-specific options]
                    189: @end example
                    190: 
                    191: The initialization options must come before the rest of the command
                    192: line. They are:
                    193: 
                    194: @table @code
                    195: @item --image-file @var{file}
                    196: Loads the Forth image @var{file} instead of the default
                    197: @file{gforth.fi}.
                    198: 
                    199: @item --path @var{path}
                    200: Uses @var{path} for searching the image file and Forth source code
                    201: files instead of the default in the environment variable
                    202: @code{GFORTHPATH} or the path specified at installation time (typically
                    203: @file{/usr/local/lib/gforth:.}). A path is given as a @code{:}-separated
                    204: list.
                    205: 
                    206: @item --dictionary-size @var{size}
                    207: @item -m @var{size}
                    208: Allocate @var{size} space for the Forth dictionary space instead of
                    209: using the default specified in the image (typically 256K). The
                    210: @var{size} specification consists of an integer and a unit (e.g.,
                    211: @code{4M}). The unit can be one of @code{b} (bytes), @code{e} (element
                    212: size, in this case Cells), @code{k} (kilobytes), and @code{M}
                    213: (Megabytes). If no unit is specified, @code{e} is used.
                    214: 
                    215: @item --data-stack-size @var{size}
                    216: @item -d @var{size}
                    217: Allocate @var{size} space for the data stack instead of using the
                    218: default specified in the image (typically 16K).
                    219: 
                    220: @item --return-stack-size @var{size}
                    221: @item -r @var{size}
                    222: Allocate @var{size} space for the return stack instead of using the
                    223: default specified in the image (typically 16K).
                    224: 
                    225: @item --fp-stack-size @var{size}
                    226: @item -f @var{size}
                    227: Allocate @var{size} space for the floating point stack instead of
                    228: using the default specified in the image (typically 16K). In this case
                    229: the unit specifier @code{e} refers to floating point numbers.
                    230: 
                    231: @item --locals-stack-size @var{size}
                    232: @item -l @var{size}
                    233: Allocate @var{size} space for the locals stack instead of using the
                    234: default specified in the image (typically 16K).
                    235: 
                    236: @end table
                    237: 
                    238: As explained above, the image-specific command-line arguments for the
                    239: default image @file{gforth.fi} consist of a sequence of filenames and
                    240: @code{-e @var{forth-code}} options that are interpreted in the seqence
                    241: in which they are given. The @code{-e @var{forth-code}} or
                    242: @code{--evaluate @var{forth-code}} option evaluates the forth
                    243: code. This option takes only one argument; if you want to evaluate more
                    244: Forth words, you have to quote them or use several @code{-e}s. To exit
                    245: after processing the command line (instead of entering interactive mode)
                    246: append @code{-e bye} to the command line.
                    247: 
                    248: Not yet implemented:
                    249: On startup the system first executes the system initialization file
                    250: (unless the option @code{--no-init-file} is given; note that the system
                    251: resulting from using this option may not be ANS Forth conformant). Then
                    252: the user initialization file @file{.gforth.fs} is executed, unless the
                    253: option @code{--no-rc} is given; this file is first searched in @file{.},
                    254: then in @file{~}, then in the normal path (see above).
                    255: 
1.4       anton     256: @node Words, ANS conformance, Invocation, Top
1.1       anton     257: @chapter Forth Words
                    258: 
                    259: @menu
1.4       anton     260: * Notation::                    
                    261: * Arithmetic::                  
                    262: * Stack Manipulation::          
                    263: * Memory access::               
                    264: * Control Structures::          
                    265: * Locals::                      
                    266: * Defining Words::              
                    267: * Wordlists::                   
                    268: * Files::                       
                    269: * Blocks::                      
                    270: * Other I/O::                   
                    271: * Programming Tools::           
1.18    ! anton     272: * Assembler and Code words::    
1.4       anton     273: * Threading Words::             
1.1       anton     274: @end menu
                    275: 
                    276: @node Notation, Arithmetic, Words, Words
                    277: @section Notation
                    278: 
                    279: The Forth words are described in this section in the glossary notation
                    280: that has become a de-facto standard for Forth texts, i.e.
                    281: 
1.4       anton     282: @format
1.1       anton     283: @var{word}     @var{Stack effect}   @var{wordset}   @var{pronunciation}
1.4       anton     284: @end format
1.1       anton     285: @var{Description}
                    286: 
                    287: @table @var
                    288: @item word
1.17      anton     289: The name of the word. BTW, Gforth is case insensitive, so you can
1.14      anton     290: type the words in in lower case (However, @pxref{core-idef}).
1.1       anton     291: 
                    292: @item Stack effect
                    293: The stack effect is written in the notation @code{@var{before} --
                    294: @var{after}}, where @var{before} and @var{after} describe the top of
                    295: stack entries before and after the execution of the word. The rest of
                    296: the stack is not touched by the word. The top of stack is rightmost,
1.17      anton     297: i.e., a stack sequence is written as it is typed in. Note that Gforth
1.1       anton     298: uses a separate floating point stack, but a unified stack
                    299: notation. Also, return stack effects are not shown in @var{stack
                    300: effect}, but in @var{Description}. The name of a stack item describes
                    301: the type and/or the function of the item. See below for a discussion of
                    302: the types.
                    303: 
                    304: @item pronunciation
                    305: How the word is pronounced
                    306: 
                    307: @item wordset
                    308: The ANS Forth standard is divided into several wordsets. A standard
                    309: system need not support all of them. So, the fewer wordsets your program
                    310: uses the more portable it will be in theory. However, we suspect that
                    311: most ANS Forth systems on personal machines will feature all
                    312: wordsets. Words that are not defined in the ANS standard have
                    313: @code{gforth} as wordset.
                    314: 
                    315: @item Description
                    316: A description of the behaviour of the word.
                    317: @end table
                    318: 
1.4       anton     319: The type of a stack item is specified by the character(s) the name
                    320: starts with:
1.1       anton     321: 
                    322: @table @code
                    323: @item f
                    324: Bool, i.e. @code{false} or @code{true}.
                    325: @item c
                    326: Char
                    327: @item w
                    328: Cell, can contain an integer or an address
                    329: @item n
                    330: signed integer
                    331: @item u
                    332: unsigned integer
                    333: @item d
                    334: double sized signed integer
                    335: @item ud
                    336: double sized unsigned integer
                    337: @item r
                    338: Float
                    339: @item a_
                    340: Cell-aligned address
                    341: @item c_
                    342: Char-aligned address (note that a Char is two bytes in Windows NT)
                    343: @item f_
                    344: Float-aligned address
                    345: @item df_
                    346: Address aligned for IEEE double precision float
                    347: @item sf_
                    348: Address aligned for IEEE single precision float
                    349: @item xt
                    350: Execution token, same size as Cell
                    351: @item wid
                    352: Wordlist ID, same size as Cell
                    353: @item f83name
                    354: Pointer to a name structure
                    355: @end table
                    356: 
1.4       anton     357: @node Arithmetic, Stack Manipulation, Notation, Words
1.1       anton     358: @section Arithmetic
                    359: Forth arithmetic is not checked, i.e., you will not hear about integer
                    360: overflow on addition or multiplication, you may hear about division by
                    361: zero if you are lucky. The operator is written after the operands, but
                    362: the operands are still in the original order. I.e., the infix @code{2-1}
                    363: corresponds to @code{2 1 -}. Forth offers a variety of division
                    364: operators. If you perform division with potentially negative operands,
                    365: you do not want to use @code{/} or @code{/mod} with its undefined
                    366: behaviour, but rather @code{fm/mod} or @code{sm/mod} (probably the
1.4       anton     367: former, @pxref{Mixed precision}).
                    368: 
                    369: @menu
                    370: * Single precision::            
                    371: * Bitwise operations::          
                    372: * Mixed precision::             operations with single and double-cell integers
                    373: * Double precision::            Double-cell integer arithmetic
                    374: * Floating Point::              
                    375: @end menu
1.1       anton     376: 
1.4       anton     377: @node Single precision, Bitwise operations, Arithmetic, Arithmetic
1.1       anton     378: @subsection Single precision
                    379: doc-+
                    380: doc--
                    381: doc-*
                    382: doc-/
                    383: doc-mod
                    384: doc-/mod
                    385: doc-negate
                    386: doc-abs
                    387: doc-min
                    388: doc-max
                    389: 
1.4       anton     390: @node Bitwise operations, Mixed precision, Single precision, Arithmetic
1.1       anton     391: @subsection Bitwise operations
                    392: doc-and
                    393: doc-or
                    394: doc-xor
                    395: doc-invert
                    396: doc-2*
                    397: doc-2/
                    398: 
1.4       anton     399: @node Mixed precision, Double precision, Bitwise operations, Arithmetic
1.1       anton     400: @subsection Mixed precision
                    401: doc-m+
                    402: doc-*/
                    403: doc-*/mod
                    404: doc-m*
                    405: doc-um*
                    406: doc-m*/
                    407: doc-um/mod
                    408: doc-fm/mod
                    409: doc-sm/rem
                    410: 
1.4       anton     411: @node Double precision, Floating Point, Mixed precision, Arithmetic
1.1       anton     412: @subsection Double precision
1.16      anton     413: 
                    414: The outer (aka text) interpreter converts numbers containing a dot into
                    415: a double precision number. Note that only numbers with the dot as last
                    416: character are standard-conforming.
                    417: 
1.1       anton     418: doc-d+
                    419: doc-d-
                    420: doc-dnegate
                    421: doc-dabs
                    422: doc-dmin
                    423: doc-dmax
                    424: 
1.4       anton     425: @node Floating Point,  , Double precision, Arithmetic
                    426: @subsection Floating Point
1.16      anton     427: 
                    428: The format of floating point numbers recognized by the outer (aka text)
                    429: interpreter is: a signed decimal number, possibly containing a decimal
                    430: point (@code{.}), followed by @code{E} or @code{e}, optionally followed
                    431: by a signed integer (the exponent). E.g., @code{1e} ist the same as
                    432: @code{+1.0e+1}. Note that a number without @code{e}
                    433: is not interpreted as floating-point number, but as double (if the
                    434: number contains a @code{.}) or single precision integer. Also,
                    435: conversions between string and floating point numbers always use base
                    436: 10, irrespective of the value of @code{BASE}. If @code{BASE} contains a
                    437: value greater then 14, the @code{E} may be interpreted as digit and the
                    438: number will be interpreted as integer, unless it has a signed exponent
                    439: (both @code{+} and @code{-} are allowed as signs).
1.4       anton     440: 
                    441: Angles in floating point operations are given in radians (a full circle
1.17      anton     442: has 2 pi radians). Note, that Gforth has a separate floating point
1.4       anton     443: stack, but we use the unified notation.
                    444: 
                    445: Floating point numbers have a number of unpleasant surprises for the
                    446: unwary (e.g., floating point addition is not associative) and even a few
                    447: for the wary. You should not use them unless you know what you are doing
                    448: or you don't care that the results you get are totally bogus. If you
                    449: want to learn about the problems of floating point numbers (and how to
1.11      anton     450: avoid them), you might start with @cite{David Goldberg, What Every
1.6       anton     451: Computer Scientist Should Know About Floating-Point Arithmetic, ACM
                    452: Computing Surveys 23(1):5@minus{}48, March 1991}.
1.4       anton     453: 
                    454: doc-f+
                    455: doc-f-
                    456: doc-f*
                    457: doc-f/
                    458: doc-fnegate
                    459: doc-fabs
                    460: doc-fmax
                    461: doc-fmin
                    462: doc-floor
                    463: doc-fround
                    464: doc-f**
                    465: doc-fsqrt
                    466: doc-fexp
                    467: doc-fexpm1
                    468: doc-fln
                    469: doc-flnp1
                    470: doc-flog
1.6       anton     471: doc-falog
1.4       anton     472: doc-fsin
                    473: doc-fcos
                    474: doc-fsincos
                    475: doc-ftan
                    476: doc-fasin
                    477: doc-facos
                    478: doc-fatan
                    479: doc-fatan2
                    480: doc-fsinh
                    481: doc-fcosh
                    482: doc-ftanh
                    483: doc-fasinh
                    484: doc-facosh
                    485: doc-fatanh
                    486: 
                    487: @node Stack Manipulation, Memory access, Arithmetic, Words
1.1       anton     488: @section Stack Manipulation
                    489: 
1.17      anton     490: Gforth has a data stack (aka parameter stack) for characters, cells,
1.1       anton     491: addresses, and double cells, a floating point stack for floating point
                    492: numbers, a return stack for storing the return addresses of colon
                    493: definitions and other data, and a locals stack for storing local
                    494: variables. Note that while every sane Forth has a separate floating
                    495: point stack, this is not strictly required; an ANS Forth system could
                    496: theoretically keep floating point numbers on the data stack. As an
                    497: additional difficulty, you don't know how many cells a floating point
                    498: number takes. It is reportedly possible to write words in a way that
                    499: they work also for a unified stack model, but we do not recommend trying
1.4       anton     500: it. Instead, just say that your program has an environmental dependency
                    501: on a separate FP stack.
                    502: 
                    503: Also, a Forth system is allowed to keep the local variables on the
1.1       anton     504: return stack. This is reasonable, as local variables usually eliminate
                    505: the need to use the return stack explicitly. So, if you want to produce
                    506: a standard complying program and if you are using local variables in a
                    507: word, forget about return stack manipulations in that word (see the
                    508: standard document for the exact rules).
                    509: 
1.4       anton     510: @menu
                    511: * Data stack::                  
                    512: * Floating point stack::        
                    513: * Return stack::                
                    514: * Locals stack::                
                    515: * Stack pointer manipulation::  
                    516: @end menu
                    517: 
                    518: @node Data stack, Floating point stack, Stack Manipulation, Stack Manipulation
1.1       anton     519: @subsection Data stack
                    520: doc-drop
                    521: doc-nip
                    522: doc-dup
                    523: doc-over
                    524: doc-tuck
                    525: doc-swap
                    526: doc-rot
                    527: doc--rot
                    528: doc-?dup
                    529: doc-pick
                    530: doc-roll
                    531: doc-2drop
                    532: doc-2nip
                    533: doc-2dup
                    534: doc-2over
                    535: doc-2tuck
                    536: doc-2swap
                    537: doc-2rot
                    538: 
1.4       anton     539: @node Floating point stack, Return stack, Data stack, Stack Manipulation
1.1       anton     540: @subsection Floating point stack
                    541: doc-fdrop
                    542: doc-fnip
                    543: doc-fdup
                    544: doc-fover
                    545: doc-ftuck
                    546: doc-fswap
                    547: doc-frot
                    548: 
1.4       anton     549: @node Return stack, Locals stack, Floating point stack, Stack Manipulation
1.1       anton     550: @subsection Return stack
                    551: doc->r
                    552: doc-r>
                    553: doc-r@
                    554: doc-rdrop
                    555: doc-2>r
                    556: doc-2r>
                    557: doc-2r@
                    558: doc-2rdrop
                    559: 
1.4       anton     560: @node Locals stack, Stack pointer manipulation, Return stack, Stack Manipulation
1.1       anton     561: @subsection Locals stack
                    562: 
1.4       anton     563: @node Stack pointer manipulation,  , Locals stack, Stack Manipulation
1.1       anton     564: @subsection Stack pointer manipulation
                    565: doc-sp@
                    566: doc-sp!
                    567: doc-fp@
                    568: doc-fp!
                    569: doc-rp@
                    570: doc-rp!
                    571: doc-lp@
                    572: doc-lp!
                    573: 
1.4       anton     574: @node Memory access, Control Structures, Stack Manipulation, Words
1.1       anton     575: @section Memory access
                    576: 
1.4       anton     577: @menu
                    578: * Stack-Memory transfers::      
                    579: * Address arithmetic::          
                    580: * Memory block access::         
                    581: @end menu
                    582: 
                    583: @node Stack-Memory transfers, Address arithmetic, Memory access, Memory access
1.1       anton     584: @subsection Stack-Memory transfers
                    585: 
                    586: doc-@
                    587: doc-!
                    588: doc-+!
                    589: doc-c@
                    590: doc-c!
                    591: doc-2@
                    592: doc-2!
                    593: doc-f@
                    594: doc-f!
                    595: doc-sf@
                    596: doc-sf!
                    597: doc-df@
                    598: doc-df!
                    599: 
1.4       anton     600: @node Address arithmetic, Memory block access, Stack-Memory transfers, Memory access
1.1       anton     601: @subsection Address arithmetic
                    602: 
                    603: ANS Forth does not specify the sizes of the data types. Instead, it
                    604: offers a number of words for computing sizes and doing address
                    605: arithmetic. Basically, address arithmetic is performed in terms of
                    606: address units (aus); on most systems the address unit is one byte. Note
                    607: that a character may have more than one au, so @code{chars} is no noop
                    608: (on systems where it is a noop, it compiles to nothing).
                    609: 
                    610: ANS Forth also defines words for aligning addresses for specific
                    611: addresses. Many computers require that accesses to specific data types
                    612: must only occur at specific addresses; e.g., that cells may only be
                    613: accessed at addresses divisible by 4. Even if a machine allows unaligned
                    614: accesses, it can usually perform aligned accesses faster. 
                    615: 
1.17      anton     616: For the performance-conscious: alignment operations are usually only
1.1       anton     617: necessary during the definition of a data structure, not during the
                    618: (more frequent) accesses to it.
                    619: 
                    620: ANS Forth defines no words for character-aligning addresses. This is not
                    621: an oversight, but reflects the fact that addresses that are not
                    622: char-aligned have no use in the standard and therefore will not be
                    623: created.
                    624: 
                    625: The standard guarantees that addresses returned by @code{CREATE}d words
1.17      anton     626: are cell-aligned; in addition, Gforth guarantees that these addresses
1.1       anton     627: are aligned for all purposes.
                    628: 
1.9       anton     629: Note that the standard defines a word @code{char}, which has nothing to
                    630: do with address arithmetic.
                    631: 
1.1       anton     632: doc-chars
                    633: doc-char+
                    634: doc-cells
                    635: doc-cell+
                    636: doc-align
                    637: doc-aligned
                    638: doc-floats
                    639: doc-float+
                    640: doc-falign
                    641: doc-faligned
                    642: doc-sfloats
                    643: doc-sfloat+
                    644: doc-sfalign
                    645: doc-sfaligned
                    646: doc-dfloats
                    647: doc-dfloat+
                    648: doc-dfalign
                    649: doc-dfaligned
1.10      anton     650: doc-maxalign
                    651: doc-maxaligned
                    652: doc-cfalign
                    653: doc-cfaligned
1.1       anton     654: doc-address-unit-bits
                    655: 
1.4       anton     656: @node Memory block access,  , Address arithmetic, Memory access
1.1       anton     657: @subsection Memory block access
                    658: 
                    659: doc-move
                    660: doc-erase
                    661: 
                    662: While the previous words work on address units, the rest works on
                    663: characters.
                    664: 
                    665: doc-cmove
                    666: doc-cmove>
                    667: doc-fill
                    668: doc-blank
                    669: 
1.4       anton     670: @node Control Structures, Locals, Memory access, Words
1.1       anton     671: @section Control Structures
                    672: 
                    673: Control structures in Forth cannot be used in interpret state, only in
                    674: compile state, i.e., in a colon definition. We do not like this
                    675: limitation, but have not seen a satisfying way around it yet, although
                    676: many schemes have been proposed.
                    677: 
1.4       anton     678: @menu
                    679: * Selection::                   
                    680: * Simple Loops::                
                    681: * Counted Loops::               
                    682: * Arbitrary control structures::  
                    683: * Calls and returns::           
                    684: * Exception Handling::          
                    685: @end menu
                    686: 
                    687: @node Selection, Simple Loops, Control Structures, Control Structures
1.1       anton     688: @subsection Selection
                    689: 
                    690: @example
                    691: @var{flag}
                    692: IF
                    693:   @var{code}
                    694: ENDIF
                    695: @end example
                    696: or
                    697: @example
                    698: @var{flag}
                    699: IF
                    700:   @var{code1}
                    701: ELSE
                    702:   @var{code2}
                    703: ENDIF
                    704: @end example
                    705: 
1.4       anton     706: You can use @code{THEN} instead of @code{ENDIF}. Indeed, @code{THEN} is
1.1       anton     707: standard, and @code{ENDIF} is not, although it is quite popular. We
                    708: recommend using @code{ENDIF}, because it is less confusing for people
                    709: who also know other languages (and is not prone to reinforcing negative
                    710: prejudices against Forth in these people). Adding @code{ENDIF} to a
                    711: system that only supplies @code{THEN} is simple:
                    712: @example
                    713: : endif   POSTPONE then ; immediate
                    714: @end example
                    715: 
                    716: [According to @cite{Webster's New Encyclopedic Dictionary}, @dfn{then
                    717: (adv.)}  has the following meanings:
                    718: @quotation
                    719: ... 2b: following next after in order ... 3d: as a necessary consequence
                    720: (if you were there, then you saw them).
                    721: @end quotation
                    722: Forth's @code{THEN} has the meaning 2b, whereas @code{THEN} in Pascal
                    723: and many other programming languages has the meaning 3d.]
                    724: 
                    725: We also provide the words @code{?dup-if} and @code{?dup-0=-if}, so you
                    726: can avoid using @code{?dup}.
                    727: 
                    728: @example
                    729: @var{n}
                    730: CASE
                    731:   @var{n1} OF @var{code1} ENDOF
                    732:   @var{n2} OF @var{code2} ENDOF
1.4       anton     733:   @dots{}
1.1       anton     734: ENDCASE
                    735: @end example
                    736: 
                    737: Executes the first @var{codei}, where the @var{ni} is equal to
                    738: @var{n}. A default case can be added by simply writing the code after
                    739: the last @code{ENDOF}. It may use @var{n}, which is on top of the stack,
                    740: but must not consume it.
                    741: 
1.4       anton     742: @node Simple Loops, Counted Loops, Selection, Control Structures
1.1       anton     743: @subsection Simple Loops
                    744: 
                    745: @example
                    746: BEGIN
                    747:   @var{code1}
                    748:   @var{flag}
                    749: WHILE
                    750:   @var{code2}
                    751: REPEAT
                    752: @end example
                    753: 
                    754: @var{code1} is executed and @var{flag} is computed. If it is true,
                    755: @var{code2} is executed and the loop is restarted; If @var{flag} is false, execution continues after the @code{REPEAT}.
                    756: 
                    757: @example
                    758: BEGIN
                    759:   @var{code}
                    760:   @var{flag}
                    761: UNTIL
                    762: @end example
                    763: 
                    764: @var{code} is executed. The loop is restarted if @code{flag} is false.
                    765: 
                    766: @example
                    767: BEGIN
                    768:   @var{code}
                    769: AGAIN
                    770: @end example
                    771: 
                    772: This is an endless loop.
                    773: 
1.4       anton     774: @node Counted Loops, Arbitrary control structures, Simple Loops, Control Structures
1.1       anton     775: @subsection Counted Loops
                    776: 
                    777: The basic counted loop is:
                    778: @example
                    779: @var{limit} @var{start}
                    780: ?DO
                    781:   @var{body}
                    782: LOOP
                    783: @end example
                    784: 
                    785: This performs one iteration for every integer, starting from @var{start}
                    786: and up to, but excluding @var{limit}. The counter, aka index, can be
                    787: accessed with @code{i}. E.g., the loop
                    788: @example
                    789: 10 0 ?DO
                    790:   i .
                    791: LOOP
                    792: @end example
                    793: prints
                    794: @example
                    795: 0 1 2 3 4 5 6 7 8 9
                    796: @end example
                    797: The index of the innermost loop can be accessed with @code{i}, the index
                    798: of the next loop with @code{j}, and the index of the third loop with
                    799: @code{k}.
                    800: 
                    801: The loop control data are kept on the return stack, so there are some
                    802: restrictions on mixing return stack accesses and counted loop
                    803: words. E.g., if you put values on the return stack outside the loop, you
                    804: cannot read them inside the loop. If you put values on the return stack
                    805: within a loop, you have to remove them before the end of the loop and
                    806: before accessing the index of the loop.
                    807: 
                    808: There are several variations on the counted loop:
                    809: 
                    810: @code{LEAVE} leaves the innermost counted loop immediately.
                    811: 
1.18    ! anton     812: If @var{start} is greater than @var{limit}, a @code{?DO} loop is entered
        !           813: (and @code{LOOP} iterates until they become equal by wrap-around
        !           814: arithmetic). This behaviour is usually not what you want. Therefore,
        !           815: Gforth offers @code{+DO} and @code{U+DO} (as replacements for
        !           816: @code{?DO}), which do not enter the loop if @var{start} is greater than
        !           817: @var{limit}; @code{+DO} is for signed loop parameters, @code{U+DO} for
        !           818: unsigned loop parameters. These words can be implemented easily on
        !           819: standard systems, so using them does not make your programs hard to
        !           820: port; e.g.:
        !           821: @example
        !           822: : +DO ( compile-time: -- do-sys; run-time: n1 n2 -- )
        !           823:     POSTPONE over POSTPONE min POSTPONE ?DO ; immediate
        !           824: @end example
        !           825: 
1.1       anton     826: @code{LOOP} can be replaced with @code{@var{n} +LOOP}; this updates the
                    827: index by @var{n} instead of by 1. The loop is terminated when the border
                    828: between @var{limit-1} and @var{limit} is crossed. E.g.:
                    829: 
1.18    ! anton     830: @code{4 0 +DO  i .  2 +LOOP}   prints @code{0 2}
1.1       anton     831: 
1.18    ! anton     832: @code{4 1 +DO  i .  2 +LOOP}   prints @code{1 3}
1.1       anton     833: 
                    834: The behaviour of @code{@var{n} +LOOP} is peculiar when @var{n} is negative:
                    835: 
1.2       anton     836: @code{-1 0 ?DO  i .  -1 +LOOP}  prints @code{0 -1}
1.1       anton     837: 
1.2       anton     838: @code{ 0 0 ?DO  i .  -1 +LOOP}  prints nothing
1.1       anton     839: 
1.18    ! anton     840: Therefore we recommend avoiding @code{@var{n} +LOOP} with negative
        !           841: @var{n}. One alternative is @code{@var{u} -LOOP}, which reduces the
        !           842: index by @var{u} each iteration. The loop is terminated when the border
        !           843: between @var{limit+1} and @var{limit} is crossed. Gforth also provides
        !           844: @code{-DO} and @code{U-DO} for down-counting loops. E.g.:
1.1       anton     845: 
1.18    ! anton     846: @code{-2 0 -DO  i .  1 -LOOP}  prints @code{0 -1}
1.1       anton     847: 
1.18    ! anton     848: @code{-1 0 -DO  i .  1 -LOOP}  prints @code{0}
1.1       anton     849: 
1.18    ! anton     850: @code{ 0 0 -DO  i .  1 -LOOP}  prints nothing
1.1       anton     851: 
1.18    ! anton     852: Another alternative is @code{@var{n} S+LOOP}, where the negative
        !           853: case behaves symmetrical to the positive case:
1.1       anton     854: 
1.18    ! anton     855: @code{-2 0 -DO  i .  -1 S+LOOP}  prints @code{0 -1}
        !           856: 
        !           857: The loop is terminated when the border between @var{limit@minus{}sgn(n)}
        !           858: and @var{limit} is crossed. Unfortunately, neither @code{-LOOP} nor
        !           859: @code{S+LOOP} are part of the ANS Forth standard, and they are not easy
        !           860: to implement using standard words. If you want to write standard
        !           861: programs, just avoid counting down.
        !           862: 
        !           863: @code{?DO} can also be replaced by @code{DO}. @code{DO} always enters
        !           864: the loop, independent of the loop parameters. Do not use @code{DO}, even
        !           865: if you know that the loop is entered in any case. Such knowledge tends
        !           866: to become invalid during maintenance of a program, and then the
        !           867: @code{DO} will make trouble.
1.1       anton     868: 
                    869: @code{UNLOOP} is used to prepare for an abnormal loop exit, e.g., via
                    870: @code{EXIT}. @code{UNLOOP} removes the loop control parameters from the
                    871: return stack so @code{EXIT} can get to its return address.
                    872: 
                    873: Another counted loop is
                    874: @example
                    875: @var{n}
                    876: FOR
                    877:   @var{body}
                    878: NEXT
                    879: @end example
                    880: This is the preferred loop of native code compiler writers who are too
1.17      anton     881: lazy to optimize @code{?DO} loops properly. In Gforth, this loop
1.1       anton     882: iterates @var{n+1} times; @code{i} produces values starting with @var{n}
                    883: and ending with 0. Other Forth systems may behave differently, even if
                    884: they support @code{FOR} loops.
                    885: 
1.4       anton     886: @node Arbitrary control structures, Calls and returns, Counted Loops, Control Structures
1.2       anton     887: @subsection Arbitrary control structures
                    888: 
                    889: ANS Forth permits and supports using control structures in a non-nested
                    890: way. Information about incomplete control structures is stored on the
                    891: control-flow stack. This stack may be implemented on the Forth data
1.17      anton     892: stack, and this is what we have done in Gforth.
1.2       anton     893: 
                    894: An @i{orig} entry represents an unresolved forward branch, a @i{dest}
                    895: entry represents a backward branch target. A few words are the basis for
                    896: building any control structure possible (except control structures that
                    897: need storage, like calls, coroutines, and backtracking).
                    898: 
1.3       anton     899: doc-if
                    900: doc-ahead
                    901: doc-then
                    902: doc-begin
                    903: doc-until
                    904: doc-again
                    905: doc-cs-pick
                    906: doc-cs-roll
1.2       anton     907: 
1.17      anton     908: On many systems control-flow stack items take one word, in Gforth they
1.2       anton     909: currently take three (this may change in the future). Therefore it is a
                    910: really good idea to manipulate the control flow stack with
                    911: @code{cs-pick} and @code{cs-roll}, not with data stack manipulation
                    912: words.
                    913: 
                    914: Some standard control structure words are built from these words:
                    915: 
1.3       anton     916: doc-else
                    917: doc-while
                    918: doc-repeat
1.2       anton     919: 
                    920: Counted loop words constitute a separate group of words:
                    921: 
1.3       anton     922: doc-?do
1.18    ! anton     923: doc-+do
        !           924: doc-u+do
        !           925: doc--do
        !           926: doc-u-do
1.3       anton     927: doc-do
                    928: doc-for
                    929: doc-loop
                    930: doc-s+loop
                    931: doc-+loop
1.18    ! anton     932: doc--loop
1.3       anton     933: doc-next
                    934: doc-leave
                    935: doc-?leave
                    936: doc-unloop
1.10      anton     937: doc-done
1.2       anton     938: 
                    939: The standard does not allow using @code{cs-pick} and @code{cs-roll} on
                    940: @i{do-sys}. Our system allows it, but it's your job to ensure that for
                    941: every @code{?DO} etc. there is exactly one @code{UNLOOP} on any path
1.3       anton     942: through the definition (@code{LOOP} etc. compile an @code{UNLOOP} on the
                    943: fall-through path). Also, you have to ensure that all @code{LEAVE}s are
1.7       pazsan    944: resolved (by using one of the loop-ending words or @code{DONE}).
1.2       anton     945: 
                    946: Another group of control structure words are
                    947: 
1.3       anton     948: doc-case
                    949: doc-endcase
                    950: doc-of
                    951: doc-endof
1.2       anton     952: 
                    953: @i{case-sys} and @i{of-sys} cannot be processed using @code{cs-pick} and
                    954: @code{cs-roll}.
                    955: 
1.3       anton     956: @subsubsection Programming Style
                    957: 
                    958: In order to ensure readability we recommend that you do not create
                    959: arbitrary control structures directly, but define new control structure
                    960: words for the control structure you want and use these words in your
                    961: program.
                    962: 
                    963: E.g., instead of writing
                    964: 
                    965: @example
                    966: begin
                    967:   ...
                    968: if [ 1 cs-roll ]
                    969:   ...
                    970: again then
                    971: @end example
                    972: 
                    973: we recommend defining control structure words, e.g.,
                    974: 
                    975: @example
                    976: : while ( dest -- orig dest )
                    977:  POSTPONE if
                    978:  1 cs-roll ; immediate
                    979: 
                    980: : repeat ( orig dest -- )
                    981:  POSTPONE again
                    982:  POSTPONE then ; immediate
                    983: @end example
                    984: 
                    985: and then using these to create the control structure:
                    986: 
                    987: @example
                    988: begin
                    989:   ...
                    990: while
                    991:   ...
                    992: repeat
                    993: @end example
                    994: 
                    995: That's much easier to read, isn't it? Of course, @code{BEGIN} and
                    996: @code{WHILE} are predefined, so in this example it would not be
                    997: necessary to define them.
                    998: 
1.4       anton     999: @node Calls and returns, Exception Handling, Arbitrary control structures, Control Structures
1.3       anton    1000: @subsection Calls and returns
                   1001: 
                   1002: A definition can be called simply be writing the name of the
1.17      anton    1003: definition. When the end of the definition is reached, it returns. An
                   1004: earlier return can be forced using
1.3       anton    1005: 
                   1006: doc-exit
                   1007: 
                   1008: Don't forget to clean up the return stack and @code{UNLOOP} any
                   1009: outstanding @code{?DO}...@code{LOOP}s before @code{EXIT}ing. The
                   1010: primitive compiled by @code{EXIT} is
                   1011: 
                   1012: doc-;s
                   1013: 
1.4       anton    1014: @node Exception Handling,  , Calls and returns, Control Structures
1.3       anton    1015: @subsection Exception Handling
                   1016: 
                   1017: doc-catch
                   1018: doc-throw
                   1019: 
1.4       anton    1020: @node Locals, Defining Words, Control Structures, Words
1.1       anton    1021: @section Locals
                   1022: 
1.2       anton    1023: Local variables can make Forth programming more enjoyable and Forth
                   1024: programs easier to read. Unfortunately, the locals of ANS Forth are
                   1025: laden with restrictions. Therefore, we provide not only the ANS Forth
                   1026: locals wordset, but also our own, more powerful locals wordset (we
                   1027: implemented the ANS Forth locals wordset through our locals wordset).
                   1028: 
                   1029: @menu
1.17      anton    1030: * Gforth locals::               
1.4       anton    1031: * ANS Forth locals::            
1.2       anton    1032: @end menu
                   1033: 
1.17      anton    1034: @node Gforth locals, ANS Forth locals, Locals, Locals
                   1035: @subsection Gforth locals
1.2       anton    1036: 
                   1037: Locals can be defined with
                   1038: 
                   1039: @example
                   1040: @{ local1 local2 ... -- comment @}
                   1041: @end example
                   1042: or
                   1043: @example
                   1044: @{ local1 local2 ... @}
                   1045: @end example
                   1046: 
                   1047: E.g.,
                   1048: @example
                   1049: : max @{ n1 n2 -- n3 @}
                   1050:  n1 n2 > if
                   1051:    n1
                   1052:  else
                   1053:    n2
                   1054:  endif ;
                   1055: @end example
                   1056: 
                   1057: The similarity of locals definitions with stack comments is intended. A
                   1058: locals definition often replaces the stack comment of a word. The order
                   1059: of the locals corresponds to the order in a stack comment and everything
                   1060: after the @code{--} is really a comment.
                   1061: 
                   1062: This similarity has one disadvantage: It is too easy to confuse locals
                   1063: declarations with stack comments, causing bugs and making them hard to
                   1064: find. However, this problem can be avoided by appropriate coding
                   1065: conventions: Do not use both notations in the same program. If you do,
                   1066: they should be distinguished using additional means, e.g. by position.
                   1067: 
                   1068: The name of the local may be preceded by a type specifier, e.g.,
                   1069: @code{F:} for a floating point value:
                   1070: 
                   1071: @example
                   1072: : CX* @{ F: Ar F: Ai F: Br F: Bi -- Cr Ci @}
                   1073: \ complex multiplication
                   1074:  Ar Br f* Ai Bi f* f-
                   1075:  Ar Bi f* Ai Br f* f+ ;
                   1076: @end example
                   1077: 
1.17      anton    1078: Gforth currently supports cells (@code{W:}, @code{W^}), doubles
1.2       anton    1079: (@code{D:}, @code{D^}), floats (@code{F:}, @code{F^}) and characters
                   1080: (@code{C:}, @code{C^}) in two flavours: a value-flavoured local (defined
                   1081: with @code{W:}, @code{D:} etc.) produces its value and can be changed
                   1082: with @code{TO}. A variable-flavoured local (defined with @code{W^} etc.)
                   1083: produces its address (which becomes invalid when the variable's scope is
                   1084: left). E.g., the standard word @code{emit} can be defined in therms of
                   1085: @code{type} like this:
                   1086: 
                   1087: @example
                   1088: : emit @{ C^ char* -- @}
                   1089:     char* 1 type ;
                   1090: @end example
                   1091: 
                   1092: A local without type specifier is a @code{W:} local. Both flavours of
                   1093: locals are initialized with values from the data or FP stack.
                   1094: 
                   1095: Currently there is no way to define locals with user-defined data
                   1096: structures, but we are working on it.
                   1097: 
1.17      anton    1098: Gforth allows defining locals everywhere in a colon definition. This
1.7       pazsan   1099: poses the following questions:
1.2       anton    1100: 
1.4       anton    1101: @menu
                   1102: * Where are locals visible by name?::  
1.14      anton    1103: * How long do locals live?::    
1.4       anton    1104: * Programming Style::           
                   1105: * Implementation::              
                   1106: @end menu
                   1107: 
1.17      anton    1108: @node Where are locals visible by name?, How long do locals live?, Gforth locals, Gforth locals
1.2       anton    1109: @subsubsection Where are locals visible by name?
                   1110: 
                   1111: Basically, the answer is that locals are visible where you would expect
                   1112: it in block-structured languages, and sometimes a little longer. If you
                   1113: want to restrict the scope of a local, enclose its definition in
                   1114: @code{SCOPE}...@code{ENDSCOPE}.
                   1115: 
                   1116: doc-scope
                   1117: doc-endscope
                   1118: 
                   1119: These words behave like control structure words, so you can use them
                   1120: with @code{CS-PICK} and @code{CS-ROLL} to restrict the scope in
                   1121: arbitrary ways.
                   1122: 
                   1123: If you want a more exact answer to the visibility question, here's the
                   1124: basic principle: A local is visible in all places that can only be
                   1125: reached through the definition of the local@footnote{In compiler
                   1126: construction terminology, all places dominated by the definition of the
                   1127: local.}. In other words, it is not visible in places that can be reached
                   1128: without going through the definition of the local. E.g., locals defined
                   1129: in @code{IF}...@code{ENDIF} are visible until the @code{ENDIF}, locals
                   1130: defined in @code{BEGIN}...@code{UNTIL} are visible after the
                   1131: @code{UNTIL} (until, e.g., a subsequent @code{ENDSCOPE}).
                   1132: 
                   1133: The reasoning behind this solution is: We want to have the locals
                   1134: visible as long as it is meaningful. The user can always make the
                   1135: visibility shorter by using explicit scoping. In a place that can
                   1136: only be reached through the definition of a local, the meaning of a
                   1137: local name is clear. In other places it is not: How is the local
                   1138: initialized at the control flow path that does not contain the
                   1139: definition? Which local is meant, if the same name is defined twice in
                   1140: two independent control flow paths?
                   1141: 
                   1142: This should be enough detail for nearly all users, so you can skip the
                   1143: rest of this section. If you relly must know all the gory details and
                   1144: options, read on.
                   1145: 
                   1146: In order to implement this rule, the compiler has to know which places
                   1147: are unreachable. It knows this automatically after @code{AHEAD},
                   1148: @code{AGAIN}, @code{EXIT} and @code{LEAVE}; in other cases (e.g., after
                   1149: most @code{THROW}s), you can use the word @code{UNREACHABLE} to tell the
                   1150: compiler that the control flow never reaches that place. If
                   1151: @code{UNREACHABLE} is not used where it could, the only consequence is
                   1152: that the visibility of some locals is more limited than the rule above
                   1153: says. If @code{UNREACHABLE} is used where it should not (i.e., if you
                   1154: lie to the compiler), buggy code will be produced.
                   1155: 
                   1156: Another problem with this rule is that at @code{BEGIN}, the compiler
1.3       anton    1157: does not know which locals will be visible on the incoming
                   1158: back-edge. All problems discussed in the following are due to this
                   1159: ignorance of the compiler (we discuss the problems using @code{BEGIN}
                   1160: loops as examples; the discussion also applies to @code{?DO} and other
1.2       anton    1161: loops). Perhaps the most insidious example is:
                   1162: @example
                   1163: AHEAD
                   1164: BEGIN
                   1165:   x
                   1166: [ 1 CS-ROLL ] THEN
1.4       anton    1167:   @{ x @}
1.2       anton    1168:   ...
                   1169: UNTIL
                   1170: @end example
                   1171: 
                   1172: This should be legal according to the visibility rule. The use of
                   1173: @code{x} can only be reached through the definition; but that appears
                   1174: textually below the use.
                   1175: 
                   1176: From this example it is clear that the visibility rules cannot be fully
                   1177: implemented without major headaches. Our implementation treats common
                   1178: cases as advertised and the exceptions are treated in a safe way: The
                   1179: compiler makes a reasonable guess about the locals visible after a
                   1180: @code{BEGIN}; if it is too pessimistic, the
                   1181: user will get a spurious error about the local not being defined; if the
                   1182: compiler is too optimistic, it will notice this later and issue a
                   1183: warning. In the case above the compiler would complain about @code{x}
                   1184: being undefined at its use. You can see from the obscure examples in
                   1185: this section that it takes quite unusual control structures to get the
                   1186: compiler into trouble, and even then it will often do fine.
                   1187: 
                   1188: If the @code{BEGIN} is reachable from above, the most optimistic guess
                   1189: is that all locals visible before the @code{BEGIN} will also be
                   1190: visible after the @code{BEGIN}. This guess is valid for all loops that
                   1191: are entered only through the @code{BEGIN}, in particular, for normal
                   1192: @code{BEGIN}...@code{WHILE}...@code{REPEAT} and
                   1193: @code{BEGIN}...@code{UNTIL} loops and it is implemented in our
                   1194: compiler. When the branch to the @code{BEGIN} is finally generated by
                   1195: @code{AGAIN} or @code{UNTIL}, the compiler checks the guess and
                   1196: warns the user if it was too optimisitic:
                   1197: @example
                   1198: IF
1.4       anton    1199:   @{ x @}
1.2       anton    1200: BEGIN
                   1201:   \ x ? 
                   1202: [ 1 cs-roll ] THEN
                   1203:   ...
                   1204: UNTIL
                   1205: @end example
                   1206: 
                   1207: Here, @code{x} lives only until the @code{BEGIN}, but the compiler
                   1208: optimistically assumes that it lives until the @code{THEN}. It notices
                   1209: this difference when it compiles the @code{UNTIL} and issues a
                   1210: warning. The user can avoid the warning, and make sure that @code{x}
                   1211: is not used in the wrong area by using explicit scoping:
                   1212: @example
                   1213: IF
                   1214:   SCOPE
1.4       anton    1215:   @{ x @}
1.2       anton    1216:   ENDSCOPE
                   1217: BEGIN
                   1218: [ 1 cs-roll ] THEN
                   1219:   ...
                   1220: UNTIL
                   1221: @end example
                   1222: 
                   1223: Since the guess is optimistic, there will be no spurious error messages
                   1224: about undefined locals.
                   1225: 
                   1226: If the @code{BEGIN} is not reachable from above (e.g., after
                   1227: @code{AHEAD} or @code{EXIT}), the compiler cannot even make an
                   1228: optimistic guess, as the locals visible after the @code{BEGIN} may be
                   1229: defined later. Therefore, the compiler assumes that no locals are
1.17      anton    1230: visible after the @code{BEGIN}. However, the user can use
1.2       anton    1231: @code{ASSUME-LIVE} to make the compiler assume that the same locals are
1.17      anton    1232: visible at the BEGIN as at the point where the top control-flow stack
                   1233: item was created.
1.2       anton    1234: 
                   1235: doc-assume-live
                   1236: 
                   1237: E.g.,
                   1238: @example
1.4       anton    1239: @{ x @}
1.2       anton    1240: AHEAD
                   1241: ASSUME-LIVE
                   1242: BEGIN
                   1243:   x
                   1244: [ 1 CS-ROLL ] THEN
                   1245:   ...
                   1246: UNTIL
                   1247: @end example
                   1248: 
                   1249: Other cases where the locals are defined before the @code{BEGIN} can be
                   1250: handled by inserting an appropriate @code{CS-ROLL} before the
                   1251: @code{ASSUME-LIVE} (and changing the control-flow stack manipulation
                   1252: behind the @code{ASSUME-LIVE}).
                   1253: 
                   1254: Cases where locals are defined after the @code{BEGIN} (but should be
                   1255: visible immediately after the @code{BEGIN}) can only be handled by
                   1256: rearranging the loop. E.g., the ``most insidious'' example above can be
                   1257: arranged into:
                   1258: @example
                   1259: BEGIN
1.4       anton    1260:   @{ x @}
1.2       anton    1261:   ... 0=
                   1262: WHILE
                   1263:   x
                   1264: REPEAT
                   1265: @end example
                   1266: 
1.17      anton    1267: @node How long do locals live?, Programming Style, Where are locals visible by name?, Gforth locals
1.2       anton    1268: @subsubsection How long do locals live?
                   1269: 
                   1270: The right answer for the lifetime question would be: A local lives at
                   1271: least as long as it can be accessed. For a value-flavoured local this
                   1272: means: until the end of its visibility. However, a variable-flavoured
                   1273: local could be accessed through its address far beyond its visibility
                   1274: scope. Ultimately, this would mean that such locals would have to be
                   1275: garbage collected. Since this entails un-Forth-like implementation
                   1276: complexities, I adopted the same cowardly solution as some other
                   1277: languages (e.g., C): The local lives only as long as it is visible;
                   1278: afterwards its address is invalid (and programs that access it
                   1279: afterwards are erroneous).
                   1280: 
1.17      anton    1281: @node Programming Style, Implementation, How long do locals live?, Gforth locals
1.2       anton    1282: @subsubsection Programming Style
                   1283: 
                   1284: The freedom to define locals anywhere has the potential to change
                   1285: programming styles dramatically. In particular, the need to use the
                   1286: return stack for intermediate storage vanishes. Moreover, all stack
                   1287: manipulations (except @code{PICK}s and @code{ROLL}s with run-time
                   1288: determined arguments) can be eliminated: If the stack items are in the
                   1289: wrong order, just write a locals definition for all of them; then
                   1290: write the items in the order you want.
                   1291: 
                   1292: This seems a little far-fetched and eliminating stack manipulations is
1.4       anton    1293: unlikely to become a conscious programming objective. Still, the number
                   1294: of stack manipulations will be reduced dramatically if local variables
1.17      anton    1295: are used liberally (e.g., compare @code{max} in @ref{Gforth locals} with
1.4       anton    1296: a traditional implementation of @code{max}).
1.2       anton    1297: 
                   1298: This shows one potential benefit of locals: making Forth programs more
                   1299: readable. Of course, this benefit will only be realized if the
                   1300: programmers continue to honour the principle of factoring instead of
                   1301: using the added latitude to make the words longer.
                   1302: 
                   1303: Using @code{TO} can and should be avoided.  Without @code{TO},
                   1304: every value-flavoured local has only a single assignment and many
                   1305: advantages of functional languages apply to Forth. I.e., programs are
                   1306: easier to analyse, to optimize and to read: It is clear from the
                   1307: definition what the local stands for, it does not turn into something
                   1308: different later.
                   1309: 
                   1310: E.g., a definition using @code{TO} might look like this:
                   1311: @example
                   1312: : strcmp @{ addr1 u1 addr2 u2 -- n @}
                   1313:  u1 u2 min 0
                   1314:  ?do
                   1315:    addr1 c@ addr2 c@ - ?dup
                   1316:    if
                   1317:      unloop exit
                   1318:    then
                   1319:    addr1 char+ TO addr1
                   1320:    addr2 char+ TO addr2
                   1321:  loop
                   1322:  u1 u2 - ;
                   1323: @end example
                   1324: Here, @code{TO} is used to update @code{addr1} and @code{addr2} at
                   1325: every loop iteration. @code{strcmp} is a typical example of the
                   1326: readability problems of using @code{TO}. When you start reading
                   1327: @code{strcmp}, you think that @code{addr1} refers to the start of the
                   1328: string. Only near the end of the loop you realize that it is something
                   1329: else.
                   1330: 
                   1331: This can be avoided by defining two locals at the start of the loop that
                   1332: are initialized with the right value for the current iteration.
                   1333: @example
                   1334: : strcmp @{ addr1 u1 addr2 u2 -- n @}
                   1335:  addr1 addr2
                   1336:  u1 u2 min 0 
                   1337:  ?do @{ s1 s2 @}
                   1338:    s1 c@ s2 c@ - ?dup 
                   1339:    if
                   1340:      unloop exit
                   1341:    then
                   1342:    s1 char+ s2 char+
                   1343:  loop
                   1344:  2drop
                   1345:  u1 u2 - ;
                   1346: @end example
                   1347: Here it is clear from the start that @code{s1} has a different value
                   1348: in every loop iteration.
                   1349: 
1.17      anton    1350: @node Implementation,  , Programming Style, Gforth locals
1.2       anton    1351: @subsubsection Implementation
                   1352: 
1.17      anton    1353: Gforth uses an extra locals stack. The most compelling reason for
1.2       anton    1354: this is that the return stack is not float-aligned; using an extra stack
                   1355: also eliminates the problems and restrictions of using the return stack
                   1356: as locals stack. Like the other stacks, the locals stack grows toward
                   1357: lower addresses. A few primitives allow an efficient implementation:
                   1358: 
                   1359: doc-@local#
                   1360: doc-f@local#
                   1361: doc-laddr#
                   1362: doc-lp+!#
                   1363: doc-lp!
                   1364: doc->l
                   1365: doc-f>l
                   1366: 
                   1367: In addition to these primitives, some specializations of these
                   1368: primitives for commonly occurring inline arguments are provided for
                   1369: efficiency reasons, e.g., @code{@@local0} as specialization of
                   1370: @code{@@local#} for the inline argument 0. The following compiling words
                   1371: compile the right specialized version, or the general version, as
                   1372: appropriate:
                   1373: 
1.12      anton    1374: doc-compile-@local
                   1375: doc-compile-f@local
1.2       anton    1376: doc-compile-lp+!
                   1377: 
                   1378: Combinations of conditional branches and @code{lp+!#} like
                   1379: @code{?branch-lp+!#} (the locals pointer is only changed if the branch
                   1380: is taken) are provided for efficiency and correctness in loops.
                   1381: 
                   1382: A special area in the dictionary space is reserved for keeping the
                   1383: local variable names. @code{@{} switches the dictionary pointer to this
                   1384: area and @code{@}} switches it back and generates the locals
                   1385: initializing code. @code{W:} etc.@ are normal defining words. This
                   1386: special area is cleared at the start of every colon definition.
                   1387: 
1.17      anton    1388: A special feature of Gforth's dictionary is used to implement the
1.2       anton    1389: definition of locals without type specifiers: every wordlist (aka
                   1390: vocabulary) has its own methods for searching
1.4       anton    1391: etc. (@pxref{Wordlists}). For the present purpose we defined a wordlist
1.2       anton    1392: with a special search method: When it is searched for a word, it
                   1393: actually creates that word using @code{W:}. @code{@{} changes the search
                   1394: order to first search the wordlist containing @code{@}}, @code{W:} etc.,
                   1395: and then the wordlist for defining locals without type specifiers.
                   1396: 
                   1397: The lifetime rules support a stack discipline within a colon
                   1398: definition: The lifetime of a local is either nested with other locals
                   1399: lifetimes or it does not overlap them.
                   1400: 
                   1401: At @code{BEGIN}, @code{IF}, and @code{AHEAD} no code for locals stack
                   1402: pointer manipulation is generated. Between control structure words
                   1403: locals definitions can push locals onto the locals stack. @code{AGAIN}
                   1404: is the simplest of the other three control flow words. It has to
                   1405: restore the locals stack depth of the corresponding @code{BEGIN}
                   1406: before branching. The code looks like this:
                   1407: @format
                   1408: @code{lp+!#} current-locals-size @minus{} dest-locals-size
                   1409: @code{branch} <begin>
                   1410: @end format
                   1411: 
                   1412: @code{UNTIL} is a little more complicated: If it branches back, it
                   1413: must adjust the stack just like @code{AGAIN}. But if it falls through,
                   1414: the locals stack must not be changed. The compiler generates the
                   1415: following code:
                   1416: @format
                   1417: @code{?branch-lp+!#} <begin> current-locals-size @minus{} dest-locals-size
                   1418: @end format
                   1419: The locals stack pointer is only adjusted if the branch is taken.
                   1420: 
                   1421: @code{THEN} can produce somewhat inefficient code:
                   1422: @format
                   1423: @code{lp+!#} current-locals-size @minus{} orig-locals-size
                   1424: <orig target>:
                   1425: @code{lp+!#} orig-locals-size @minus{} new-locals-size
                   1426: @end format
                   1427: The second @code{lp+!#} adjusts the locals stack pointer from the
1.4       anton    1428: level at the @var{orig} point to the level after the @code{THEN}. The
1.2       anton    1429: first @code{lp+!#} adjusts the locals stack pointer from the current
                   1430: level to the level at the orig point, so the complete effect is an
                   1431: adjustment from the current level to the right level after the
                   1432: @code{THEN}.
                   1433: 
                   1434: In a conventional Forth implementation a dest control-flow stack entry
                   1435: is just the target address and an orig entry is just the address to be
                   1436: patched. Our locals implementation adds a wordlist to every orig or dest
                   1437: item. It is the list of locals visible (or assumed visible) at the point
                   1438: described by the entry. Our implementation also adds a tag to identify
                   1439: the kind of entry, in particular to differentiate between live and dead
                   1440: (reachable and unreachable) orig entries.
                   1441: 
                   1442: A few unusual operations have to be performed on locals wordlists:
                   1443: 
                   1444: doc-common-list
                   1445: doc-sub-list?
                   1446: doc-list-size
                   1447: 
                   1448: Several features of our locals wordlist implementation make these
                   1449: operations easy to implement: The locals wordlists are organised as
                   1450: linked lists; the tails of these lists are shared, if the lists
                   1451: contain some of the same locals; and the address of a name is greater
                   1452: than the address of the names behind it in the list.
                   1453: 
                   1454: Another important implementation detail is the variable
                   1455: @code{dead-code}. It is used by @code{BEGIN} and @code{THEN} to
                   1456: determine if they can be reached directly or only through the branch
                   1457: that they resolve. @code{dead-code} is set by @code{UNREACHABLE},
                   1458: @code{AHEAD}, @code{EXIT} etc., and cleared at the start of a colon
                   1459: definition, by @code{BEGIN} and usually by @code{THEN}.
                   1460: 
                   1461: Counted loops are similar to other loops in most respects, but
                   1462: @code{LEAVE} requires special attention: It performs basically the same
                   1463: service as @code{AHEAD}, but it does not create a control-flow stack
                   1464: entry. Therefore the information has to be stored elsewhere;
                   1465: traditionally, the information was stored in the target fields of the
                   1466: branches created by the @code{LEAVE}s, by organizing these fields into a
                   1467: linked list. Unfortunately, this clever trick does not provide enough
                   1468: space for storing our extended control flow information. Therefore, we
                   1469: introduce another stack, the leave stack. It contains the control-flow
                   1470: stack entries for all unresolved @code{LEAVE}s.
                   1471: 
                   1472: Local names are kept until the end of the colon definition, even if
                   1473: they are no longer visible in any control-flow path. In a few cases
                   1474: this may lead to increased space needs for the locals name area, but
                   1475: usually less than reclaiming this space would cost in code size.
                   1476: 
                   1477: 
1.17      anton    1478: @node ANS Forth locals,  , Gforth locals, Locals
1.2       anton    1479: @subsection ANS Forth locals
                   1480: 
                   1481: The ANS Forth locals wordset does not define a syntax for locals, but
                   1482: words that make it possible to define various syntaxes. One of the
1.17      anton    1483: possible syntaxes is a subset of the syntax we used in the Gforth locals
1.2       anton    1484: wordset, i.e.:
                   1485: 
                   1486: @example
                   1487: @{ local1 local2 ... -- comment @}
                   1488: @end example
                   1489: or
                   1490: @example
                   1491: @{ local1 local2 ... @}
                   1492: @end example
                   1493: 
                   1494: The order of the locals corresponds to the order in a stack comment. The
                   1495: restrictions are:
1.1       anton    1496: 
1.2       anton    1497: @itemize @bullet
                   1498: @item
1.17      anton    1499: Locals can only be cell-sized values (no type specifiers are allowed).
1.2       anton    1500: @item
                   1501: Locals can be defined only outside control structures.
                   1502: @item
                   1503: Locals can interfere with explicit usage of the return stack. For the
                   1504: exact (and long) rules, see the standard. If you don't use return stack
1.17      anton    1505: accessing words in a definition using locals, you will be all right. The
1.2       anton    1506: purpose of this rule is to make locals implementation on the return
                   1507: stack easier.
                   1508: @item
                   1509: The whole definition must be in one line.
                   1510: @end itemize
                   1511: 
                   1512: Locals defined in this way behave like @code{VALUE}s
1.4       anton    1513: (@xref{Values}). I.e., they are initialized from the stack. Using their
1.2       anton    1514: name produces their value. Their value can be changed using @code{TO}.
                   1515: 
1.17      anton    1516: Since this syntax is supported by Gforth directly, you need not do
1.2       anton    1517: anything to use it. If you want to port a program using this syntax to
                   1518: another ANS Forth system, use @file{anslocal.fs} to implement the syntax
                   1519: on the other system.
                   1520: 
                   1521: Note that a syntax shown in the standard, section A.13 looks
                   1522: similar, but is quite different in having the order of locals
                   1523: reversed. Beware!
                   1524: 
                   1525: The ANS Forth locals wordset itself consists of the following word
                   1526: 
                   1527: doc-(local)
                   1528: 
                   1529: The ANS Forth locals extension wordset defines a syntax, but it is so
                   1530: awful that we strongly recommend not to use it. We have implemented this
1.17      anton    1531: syntax to make porting to Gforth easy, but do not document it here. The
1.2       anton    1532: problem with this syntax is that the locals are defined in an order
                   1533: reversed with respect to the standard stack comment notation, making
                   1534: programs harder to read, and easier to misread and miswrite. The only
                   1535: merit of this syntax is that it is easy to implement using the ANS Forth
                   1536: locals wordset.
1.3       anton    1537: 
1.4       anton    1538: @node Defining Words, Wordlists, Locals, Words
                   1539: @section Defining Words
                   1540: 
1.14      anton    1541: @menu
                   1542: * Values::                      
                   1543: @end menu
                   1544: 
1.4       anton    1545: @node Values,  , Defining Words, Defining Words
                   1546: @subsection Values
                   1547: 
                   1548: @node Wordlists, Files, Defining Words, Words
                   1549: @section Wordlists
                   1550: 
                   1551: @node Files, Blocks, Wordlists, Words
                   1552: @section Files
                   1553: 
                   1554: @node Blocks, Other I/O, Files, Words
                   1555: @section Blocks
                   1556: 
                   1557: @node Other I/O, Programming Tools, Blocks, Words
                   1558: @section Other I/O
                   1559: 
1.18    ! anton    1560: @node Programming Tools, Assembler and Code words, Other I/O, Words
1.4       anton    1561: @section Programming Tools
                   1562: 
1.5       anton    1563: @menu
                   1564: * Debugging::                   Simple and quick.
                   1565: * Assertions::                  Making your programs self-checking.
                   1566: @end menu
                   1567: 
                   1568: @node Debugging, Assertions, Programming Tools, Programming Tools
1.4       anton    1569: @subsection Debugging
                   1570: 
                   1571: The simple debugging aids provided in @file{debugging.fs}
                   1572: are meant to support a different style of debugging than the
                   1573: tracing/stepping debuggers used in languages with long turn-around
                   1574: times.
                   1575: 
                   1576: A much better (faster) way in fast-compilig languages is to add
                   1577: printing code at well-selected places, let the program run, look at
                   1578: the output, see where things went wrong, add more printing code, etc.,
                   1579: until the bug is found.
                   1580: 
                   1581: The word @code{~~} is easy to insert. It just prints debugging
                   1582: information (by default the source location and the stack contents). It
                   1583: is also easy to remove (@kbd{C-x ~} in the Emacs Forth mode to
                   1584: query-replace them with nothing). The deferred words
                   1585: @code{printdebugdata} and @code{printdebugline} control the output of
                   1586: @code{~~}. The default source location output format works well with
                   1587: Emacs' compilation mode, so you can step through the program at the
1.5       anton    1588: source level using @kbd{C-x `} (the advantage over a stepping debugger
                   1589: is that you can step in any direction and you know where the crash has
                   1590: happened or where the strange data has occurred).
1.4       anton    1591: 
                   1592: Note that the default actions clobber the contents of the pictured
                   1593: numeric output string, so you should not use @code{~~}, e.g., between
                   1594: @code{<#} and @code{#>}.
                   1595: 
                   1596: doc-~~
                   1597: doc-printdebugdata
                   1598: doc-printdebugline
                   1599: 
1.5       anton    1600: @node Assertions,  , Debugging, Programming Tools
1.4       anton    1601: @subsection Assertions
                   1602: 
1.5       anton    1603: It is a good idea to make your programs self-checking, in particular, if
                   1604: you use an assumption (e.g., that a certain field of a data structure is
1.17      anton    1605: never zero) that may become wrong during maintenance. Gforth supports
1.5       anton    1606: assertions for this purpose. They are used like this:
                   1607: 
                   1608: @example
                   1609: assert( @var{flag} )
                   1610: @end example
                   1611: 
                   1612: The code between @code{assert(} and @code{)} should compute a flag, that
                   1613: should be true if everything is alright and false otherwise. It should
                   1614: not change anything else on the stack. The overall stack effect of the
                   1615: assertion is @code{( -- )}. E.g.
                   1616: 
                   1617: @example
                   1618: assert( 1 1 + 2 = ) \ what we learn in school
                   1619: assert( dup 0<> ) \ assert that the top of stack is not zero
                   1620: assert( false ) \ this code should not be reached
                   1621: @end example
                   1622: 
                   1623: The need for assertions is different at different times. During
                   1624: debugging, we want more checking, in production we sometimes care more
                   1625: for speed. Therefore, assertions can be turned off, i.e., the assertion
                   1626: becomes a comment. Depending on the importance of an assertion and the
                   1627: time it takes to check it, you may want to turn off some assertions and
1.17      anton    1628: keep others turned on. Gforth provides several levels of assertions for
1.5       anton    1629: this purpose:
                   1630: 
                   1631: doc-assert0(
                   1632: doc-assert1(
                   1633: doc-assert2(
                   1634: doc-assert3(
                   1635: doc-assert(
                   1636: doc-)
                   1637: 
                   1638: @code{Assert(} is the same as @code{assert1(}. The variable
                   1639: @code{assert-level} specifies the highest assertions that are turned
                   1640: on. I.e., at the default @code{assert-level} of one, @code{assert0(} and
                   1641: @code{assert1(} assertions perform checking, while @code{assert2(} and
                   1642: @code{assert3(} assertions are treated as comments.
                   1643: 
                   1644: Note that the @code{assert-level} is evaluated at compile-time, not at
                   1645: run-time. I.e., you cannot turn assertions on or off at run-time, you
                   1646: have to set the @code{assert-level} appropriately before compiling a
                   1647: piece of code. You can compile several pieces of code at several
                   1648: @code{assert-level}s (e.g., a trusted library at level 1 and newly
                   1649: written code at level 3).
                   1650: 
                   1651: doc-assert-level
                   1652: 
                   1653: If an assertion fails, a message compatible with Emacs' compilation mode
                   1654: is produced and the execution is aborted (currently with @code{ABORT"}.
                   1655: If there is interest, we will introduce a special throw code. But if you
                   1656: intend to @code{catch} a specific condition, using @code{throw} is
                   1657: probably more appropriate than an assertion).
                   1658: 
1.18    ! anton    1659: @node Assembler and Code words, Threading Words, Programming Tools, Words
        !          1660: @section Assembler and Code words
        !          1661: 
        !          1662: Gforth provides some words for defining primitives (words written in
        !          1663: machine code), and for defining the the machine-code equivalent of
        !          1664: @code{DOES>}-based defining words. However, the machine-independent
        !          1665: nature of Gforth poses a few problems: First of all. Gforth runs on
        !          1666: several architectures, so it can provide no standard assembler. What's
        !          1667: worse is that the register allocation not only depends on the processor,
        !          1668: but also on the gcc version and options used.
        !          1669: 
        !          1670: The words Gforth offers encapsulate some system dependences (e.g., the
        !          1671: header structure), so a system-independent assembler may be used in
        !          1672: Gforth. If you do not have an assembler, you can compile machine code
        !          1673: directly with @code{,} and @code{c,}.
        !          1674: 
        !          1675: doc-assembler
        !          1676: doc-code
        !          1677: doc-end-code
        !          1678: doc-;code
        !          1679: doc-flush-icache
        !          1680: 
        !          1681: If @code{flush-icache} does not work correctly, @code{code} words
        !          1682: etc. will not work (reliably), either.
        !          1683: 
        !          1684: These words are rarely used. Therefore they reside in @code{code.fs},
        !          1685: which is usually not loaded (except @code{flush-icache}, which is always
        !          1686: present). You can load it with @code{require code.fs}.
        !          1687: 
        !          1688: Another option for implementing normal and defining words efficiently
        !          1689: is: adding the wanted functionality to the source of Gforth. For normal
        !          1690: words you just have to edit @file{primitives}, defining words (for fast
        !          1691: defined words) probably require changes in @file{engine.c},
        !          1692: @file{kernal.fs}, @file{prims2x.fs}, and possibly @file{cross.fs}.
        !          1693: 
        !          1694: 
        !          1695: @node Threading Words,  , Assembler and Code words, Words
1.4       anton    1696: @section Threading Words
                   1697: 
                   1698: These words provide access to code addresses and other threading stuff
1.17      anton    1699: in Gforth (and, possibly, other interpretive Forths). It more or less
1.4       anton    1700: abstracts away the differences between direct and indirect threading
                   1701: (and, for direct threading, the machine dependences). However, at
                   1702: present this wordset is still inclomplete. It is also pretty low-level;
                   1703: some day it will hopefully be made unnecessary by an internals words set
                   1704: that abstracts implementation details away completely.
                   1705: 
                   1706: doc->code-address
                   1707: doc->does-code
                   1708: doc-code-address!
                   1709: doc-does-code!
                   1710: doc-does-handler!
                   1711: doc-/does-handler
                   1712: 
1.18    ! anton    1713: The code addresses produced by various defining words are produced by
        !          1714: the following words:
1.14      anton    1715: 
1.18    ! anton    1716: doc-docol:
        !          1717: doc-docon:
        !          1718: doc-dovar:
        !          1719: doc-douser:
        !          1720: doc-dodefer:
        !          1721: doc-dofield:
        !          1722: 
        !          1723: Currently there is no installation-independent way for recogizing words
        !          1724: defined by a @code{CREATE}...@code{DOES>} word; however, once you know
        !          1725: that a word is defined by a @code{CREATE}...@code{DOES>} word, you can
        !          1726: use @code{>DOES-CODE}.
1.14      anton    1727: 
1.4       anton    1728: @node ANS conformance, Model, Words, Top
                   1729: @chapter ANS conformance
                   1730: 
1.17      anton    1731: To the best of our knowledge, Gforth is an
1.14      anton    1732: 
1.15      anton    1733: ANS Forth System
                   1734: @itemize
                   1735: @item providing the Core Extensions word set
                   1736: @item providing the Block word set
                   1737: @item providing the Block Extensions word set
                   1738: @item providing the Double-Number word set
                   1739: @item providing the Double-Number Extensions word set
                   1740: @item providing the Exception word set
                   1741: @item providing the Exception Extensions word set
                   1742: @item providing the Facility word set
                   1743: @item providing @code{MS} and @code{TIME&DATE} from the Facility Extensions word set
                   1744: @item providing the File Access word set
                   1745: @item providing the File Access Extensions word set
                   1746: @item providing the Floating-Point word set
                   1747: @item providing the Floating-Point Extensions word set
                   1748: @item providing the Locals word set
                   1749: @item providing the Locals Extensions word set
                   1750: @item providing the Memory-Allocation word set
                   1751: @item providing the Memory-Allocation Extensions word set (that one's easy)
                   1752: @item providing the Programming-Tools word set
1.18    ! anton    1753: @item providing @code{;code}, @code{AHEAD}, @code{ASSEMBLER}, @code{BYE}, @code{CODE}, @code{CS-PICK}, @code{CS-ROLL}, @code{STATE}, @code{[ELSE]}, @code{[IF]}, @code{[THEN]} from the Programming-Tools Extensions word set
1.15      anton    1754: @item providing the Search-Order word set
                   1755: @item providing the Search-Order Extensions word set
                   1756: @item providing the String word set
                   1757: @item providing the String Extensions word set (another easy one)
                   1758: @end itemize
                   1759: 
                   1760: In addition, ANS Forth systems are required to document certain
                   1761: implementation choices. This chapter tries to meet these
                   1762: requirements. In many cases it gives a way to ask the system for the
                   1763: information instead of providing the information directly, in
                   1764: particular, if the information depends on the processor, the operating
                   1765: system or the installation options chosen, or if they are likely to
1.17      anton    1766: change during the maintenance of Gforth.
1.15      anton    1767: 
1.14      anton    1768: @comment The framework for the rest has been taken from pfe.
                   1769: 
                   1770: @menu
                   1771: * The Core Words::              
                   1772: * The optional Block word set::  
                   1773: * The optional Double Number word set::  
                   1774: * The optional Exception word set::  
                   1775: * The optional Facility word set::  
                   1776: * The optional File-Access word set::  
                   1777: * The optional Floating-Point word set::  
                   1778: * The optional Locals word set::  
                   1779: * The optional Memory-Allocation word set::  
                   1780: * The optional Programming-Tools word set::  
                   1781: * The optional Search-Order word set::  
                   1782: @end menu
                   1783: 
                   1784: 
                   1785: @c =====================================================================
                   1786: @node The Core Words, The optional Block word set, ANS conformance, ANS conformance
                   1787: @comment  node-name,  next,  previous,  up
                   1788: @section The Core Words
                   1789: @c =====================================================================
                   1790: 
                   1791: @menu
1.15      anton    1792: * core-idef::                   Implementation Defined Options                   
                   1793: * core-ambcond::                Ambiguous Conditions                
                   1794: * core-other::                  Other System Documentation                  
1.14      anton    1795: @end menu
                   1796: 
                   1797: @c ---------------------------------------------------------------------
                   1798: @node core-idef, core-ambcond, The Core Words, The Core Words
                   1799: @subsection Implementation Defined Options
                   1800: @c ---------------------------------------------------------------------
                   1801: 
                   1802: @table @i
                   1803: 
                   1804: @item (Cell) aligned addresses:
1.17      anton    1805: processor-dependent. Gforth's alignment words perform natural alignment
1.14      anton    1806: (e.g., an address aligned for a datum of size 8 is divisible by
                   1807: 8). Unaligned accesses usually result in a @code{-23 THROW}.
                   1808: 
                   1809: @item @code{EMIT} and non-graphic characters:
                   1810: The character is output using the C library function (actually, macro)
                   1811: @code{putchar}.
                   1812: 
                   1813: @item character editing of @code{ACCEPT} and @code{EXPECT}:
                   1814: This is modeled on the GNU readline library (@pxref{Readline
                   1815: Interaction, , Command Line Editing, readline, The GNU Readline
                   1816: Library}) with Emacs-like key bindings. @kbd{Tab} deviates a little by
                   1817: producing a full word completion every time you type it (instead of
                   1818: producing the common prefix of all completions).
                   1819: 
                   1820: @item character set:
                   1821: The character set of your computer and display device. Gforth is
                   1822: 8-bit-clean (but some other component in your system may make trouble).
                   1823: 
                   1824: @item Character-aligned address requirements:
                   1825: installation-dependent. Currently a character is represented by a C
                   1826: @code{unsigned char}; in the future we might switch to @code{wchar_t}
                   1827: (Comments on that requested).
                   1828: 
                   1829: @item character-set extensions and matching of names:
1.17      anton    1830: Any character except the ASCII NUL charcter can be used in a
                   1831: name. Matching is case-insensitive. The matching is performed using the
                   1832: C function @code{strncasecmp}, whose function is probably influenced by
                   1833: the locale. E.g., the @code{C} locale does not know about accents and
1.14      anton    1834: umlauts, so they are matched case-sensitively in that locale. For
                   1835: portability reasons it is best to write programs such that they work in
                   1836: the @code{C} locale. Then one can use libraries written by a Polish
                   1837: programmer (who might use words containing ISO Latin-2 encoded
                   1838: characters) and by a French programmer (ISO Latin-1) in the same program
                   1839: (of course, @code{WORDS} will produce funny results for some of the
                   1840: words (which ones, depends on the font you are using)). Also, the locale
                   1841: you prefer may not be available in other operating systems. Hopefully,
                   1842: Unicode will solve these problems one day.
                   1843: 
                   1844: @item conditions under which control characters match a space delimiter:
                   1845: If @code{WORD} is called with the space character as a delimiter, all
                   1846: white-space characters (as identified by the C macro @code{isspace()})
                   1847: are delimiters. @code{PARSE}, on the other hand, treats space like other
                   1848: delimiters. @code{PARSE-WORD} treats space like @code{WORD}, but behaves
                   1849: like @code{PARSE} otherwise. @code{(NAME)}, which is used by the outer
                   1850: interpreter (aka text interpreter) by default, treats all white-space
                   1851: characters as delimiters.
                   1852: 
                   1853: @item format of the control flow stack:
                   1854: The data stack is used as control flow stack. The size of a control flow
                   1855: stack item in cells is given by the constant @code{cs-item-size}. At the
                   1856: time of this writing, an item consists of a (pointer to a) locals list
                   1857: (third), an address in the code (second), and a tag for identifying the
                   1858: item (TOS). The following tags are used: @code{defstart},
                   1859: @code{live-orig}, @code{dead-orig}, @code{dest}, @code{do-dest},
                   1860: @code{scopestart}.
                   1861: 
                   1862: @item conversion of digits > 35
                   1863: The characters @code{[\]^_'} are the digits with the decimal value
                   1864: 36@minus{}41. There is no way to input many of the larger digits.
                   1865: 
                   1866: @item display after input terminates in @code{ACCEPT} and @code{EXPECT}:
                   1867: The cursor is moved to the end of the entered string. If the input is
                   1868: terminated using the @kbd{Return} key, a space is typed.
                   1869: 
                   1870: @item exception abort sequence of @code{ABORT"}:
                   1871: The error string is stored into the variable @code{"error} and a
                   1872: @code{-2 throw} is performed.
                   1873: 
                   1874: @item input line terminator:
                   1875: For interactive input, @kbd{C-m} and @kbd{C-j} terminate lines. One of
                   1876: these characters is typically produced when you type the @kbd{Enter} or
                   1877: @kbd{Return} key.
                   1878: 
                   1879: @item maximum size of a counted string:
                   1880: @code{s" /counted-string" environment? drop .}. Currently 255 characters
                   1881: on all ports, but this may change.
                   1882: 
                   1883: @item maximum size of a parsed string:
                   1884: Given by the constant @code{/line}. Currently 255 characters.
                   1885: 
                   1886: @item maximum size of a definition name, in characters:
                   1887: 31
                   1888: 
                   1889: @item maximum string length for @code{ENVIRONMENT?}, in characters:
                   1890: 31
                   1891: 
                   1892: @item method of selecting the user input device:
1.17      anton    1893: The user input device is the standard input. There is currently no way to
                   1894: change it from within Gforth. However, the input can typically be
                   1895: redirected in the command line that starts Gforth.
1.14      anton    1896: 
                   1897: @item method of selecting the user output device:
                   1898: The user output device is the standard output. It cannot be redirected
1.17      anton    1899: from within Gforth, but typically from the command line that starts
                   1900: Gforth. Gforth uses buffered output, so output on a terminal does not
1.14      anton    1901: become visible before the next newline or buffer overflow. Output on
                   1902: non-terminals is invisible until the buffer overflows.
                   1903: 
                   1904: @item methods of dictionary compilation:
1.17      anton    1905: What are we expected to document here?
1.14      anton    1906: 
                   1907: @item number of bits in one address unit:
                   1908: @code{s" address-units-bits" environment? drop .}. 8 in all current
                   1909: ports.
                   1910: 
                   1911: @item number representation and arithmetic:
                   1912: Processor-dependent. Binary two's complement on all current ports.
                   1913: 
                   1914: @item ranges for integer types:
                   1915: Installation-dependent. Make environmental queries for @code{MAX-N},
                   1916: @code{MAX-U}, @code{MAX-D} and @code{MAX-UD}. The lower bounds for
                   1917: unsigned (and positive) types is 0. The lower bound for signed types on
                   1918: two's complement and one's complement machines machines can be computed
                   1919: by adding 1 to the upper bound.
                   1920: 
                   1921: @item read-only data space regions:
                   1922: The whole Forth data space is writable.
                   1923: 
                   1924: @item size of buffer at @code{WORD}:
                   1925: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
                   1926: shared with the pictured numeric output string. If overwriting
                   1927: @code{PAD} is acceptable, it is as large as the remaining dictionary
                   1928: space, although only as much can be sensibly used as fits in a counted
                   1929: string.
                   1930: 
                   1931: @item size of one cell in address units:
                   1932: @code{1 cells .}.
                   1933: 
                   1934: @item size of one character in address units:
                   1935: @code{1 chars .}. 1 on all current ports.
                   1936: 
                   1937: @item size of the keyboard terminal buffer:
                   1938: Varies. You can determine the size at a specific time using @code{lp@
                   1939: tib - .}. It is shared with the locals stack and TIBs of files that
                   1940: include the current file. You can change the amount of space for TIBs
1.17      anton    1941: and locals stack at Gforth startup with the command line option
1.14      anton    1942: @code{-l}.
                   1943: 
                   1944: @item size of the pictured numeric output buffer:
                   1945: @code{PAD HERE - .}. 104 characters on 32-bit machines. The buffer is
                   1946: shared with @code{WORD}.
                   1947: 
                   1948: @item size of the scratch area returned by @code{PAD}:
                   1949: The remainder of dictionary space. You can even use the unused part of
                   1950: the data stack space. The current size can be computed with @code{sp@
                   1951: pad - .}.
                   1952: 
                   1953: @item system case-sensitivity characteristics:
                   1954: Dictionary searches are case insensitive. However, as explained above
                   1955: under @i{character-set extensions}, the matching for non-ASCII
                   1956: characters is determined by the locale you are using. In the default
                   1957: @code{C} locale all non-ASCII characters are matched case-sensitively.
                   1958: 
                   1959: @item system prompt:
                   1960: @code{ ok} in interpret state, @code{ compiled} in compile state.
                   1961: 
                   1962: @item division rounding:
                   1963: installation dependent. @code{s" floored" environment? drop .}. We leave
                   1964: the choice to gcc (what to use for @code{/}) and to you (whether to use
                   1965: @code{fm/mod}, @code{sm/rem} or simply @code{/}).
                   1966: 
                   1967: @item values of @code{STATE} when true:
                   1968: -1.
                   1969: 
                   1970: @item values returned after arithmetic overflow:
                   1971: On two's complement machines, arithmetic is performed modulo
                   1972: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
                   1973: arithmetic (with appropriate mapping for signed types). Division by zero
                   1974: typically results in a @code{-55 throw} (floatingpoint unidentified
                   1975: fault), although a @code{-10 throw} (divide by zero) would be more
                   1976: appropriate.
                   1977: 
                   1978: @item whether the current definition can be found after @t{DOES>}:
                   1979: No.
                   1980: 
                   1981: @end table
                   1982: 
                   1983: @c ---------------------------------------------------------------------
                   1984: @node core-ambcond, core-other, core-idef, The Core Words
                   1985: @subsection Ambiguous conditions
                   1986: @c ---------------------------------------------------------------------
                   1987: 
                   1988: @table @i
                   1989: 
                   1990: @item a name is neither a word nor a number:
                   1991: @code{-13 throw} (Undefined word)
                   1992: 
                   1993: @item a definition name exceeds the maximum length allowed:
                   1994: @code{-19 throw} (Word name too long)
                   1995: 
                   1996: @item addressing a region not inside the various data spaces of the forth system:
                   1997: The stacks, code space and name space are accessible. Machine code space is
                   1998: typically readable. Accessing other addresses gives results dependent on
                   1999: the operating system. On decent systems: @code{-9 throw} (Invalid memory
                   2000: address).
                   2001: 
                   2002: @item argument type incompatible with parameter:
                   2003: This is usually not caught. Some words perform checks, e.g., the control
                   2004: flow words, and issue a @code{ABORT"} or @code{-12 THROW} (Argument type
                   2005: mismatch).
                   2006: 
                   2007: @item attempting to obtain the execution token of a word with undefined execution semantics:
                   2008: You get an execution token representing the compilation semantics
                   2009: instead.
                   2010: 
                   2011: @item dividing by zero:
                   2012: typically results in a @code{-55 throw} (floating point unidentified
                   2013: fault), although a @code{-10 throw} (divide by zero) would be more
                   2014: appropriate.
                   2015: 
                   2016: @item insufficient data stack or return stack space:
                   2017: Not checked. This typically results in mysterious illegal memory
                   2018: accesses, producing @code{-9 throw} (Invalid memory address) or
                   2019: @code{-23 throw} (Address alignment exception).
                   2020: 
                   2021: @item insufficient space for loop control parameters:
                   2022: like other return stack overflows.
                   2023: 
                   2024: @item insufficient space in the dictionary:
                   2025: Not checked. Similar results as stack overflows. However, typically the
                   2026: error appears at a different place when one inserts or removes code.
                   2027: 
                   2028: @item interpreting a word with undefined interpretation semantics:
                   2029: For some words, we defined interpretation semantics. For the others:
                   2030: @code{-14 throw} (Interpreting a compile-only word). Note that this is
                   2031: checked only by the outer (aka text) interpreter; if the word is
                   2032: @code{execute}d in some other way, it will typically perform it's
                   2033: compilation semantics even in interpret state. (We could change @code{'}
                   2034: and relatives not to give the xt of such words, but we think that would
                   2035: be too restrictive).
                   2036: 
                   2037: @item modifying the contents of the input buffer or a string literal:
                   2038: These are located in writable memory and can be modified.
                   2039: 
                   2040: @item overflow of the pictured numeric output string:
                   2041: Not checked.
                   2042: 
                   2043: @item parsed string overflow:
                   2044: @code{PARSE} cannot overflow. @code{WORD} does not check for overflow.
                   2045: 
                   2046: @item producing a result out of range:
                   2047: On two's complement machines, arithmetic is performed modulo
                   2048: 2**bits-per-cell for single arithmetic and 4**bits-per-cell for double
                   2049: arithmetic (with appropriate mapping for signed types). Division by zero
                   2050: typically results in a @code{-55 throw} (floatingpoint unidentified
                   2051: fault), although a @code{-10 throw} (divide by zero) would be more
                   2052: appropriate. @code{convert} and @code{>number} currently overflow
                   2053: silently.
                   2054: 
                   2055: @item reading from an empty data or return stack:
                   2056: The data stack is checked by the outer (aka text) interpreter after
                   2057: every word executed. If it has underflowed, a @code{-4 throw} (Stack
                   2058: underflow) is performed. Apart from that, the stacks are not checked and
                   2059: underflows can result in similar behaviour as overflows (of adjacent
                   2060: stacks).
                   2061: 
                   2062: @item unexepected end of the input buffer, resulting in an attempt to use a zero-length string as a name:
                   2063: @code{Create} and its descendants perform a @code{-16 throw} (Attempt to
                   2064: use zero-length string as a name). Words like @code{'} probably will not
                   2065: find what they search. Note that it is possible to create zero-length
                   2066: names with @code{nextname} (should it not?).
                   2067: 
                   2068: @item @code{>IN} greater than input buffer:
                   2069: The next invocation of a parsing word returns a string wih length 0.
                   2070: 
                   2071: @item @code{RECURSE} appears after @code{DOES>}:
                   2072: Compiles a recursive call to the defining word not to the defined word.
                   2073: 
                   2074: @item argument input source different than current input source for @code{RESTORE-INPUT}:
                   2075: !!???If the argument input source is a valid input source then it gets
                   2076: restored. Otherwise causes @code{-12 THROW} which unless caught issues
                   2077: the message "argument type mismatch" and aborts.
                   2078: 
                   2079: @item data space containing definitions gets de-allocated:
                   2080: Deallocation with @code{allot} is not checked. This typically resuls in
                   2081: memory access faults or execution of illegal instructions.
                   2082: 
                   2083: @item data space read/write with incorrect alignment:
                   2084: Processor-dependent. Typically results in a @code{-23 throw} (Address
                   2085: alignment exception). Under Linux on a 486 or later processor with
                   2086: alignment turned on, incorrect alignment results in a @code{-9 throw}
                   2087: (Invalid memory address). There are reportedly some processors with
                   2088: alignment restrictions that do not report them.
                   2089: 
                   2090: @item data space pointer not properly aligned, @code{,}, @code{C,}:
                   2091: Like other alignment errors.
                   2092: 
                   2093: @item less than u+2 stack items (@code{PICK} and @code{ROLL}):
                   2094: Not checked. May cause an illegal memory access.
                   2095: 
                   2096: @item loop control parameters not available:
                   2097: Not checked. The counted loop words simply assume that the top of return
                   2098: stack items are loop control parameters and behave accordingly.
                   2099: 
                   2100: @item most recent definition does not have a name (@code{IMMEDIATE}):
                   2101: @code{abort" last word was headerless"}.
                   2102: 
                   2103: @item name not defined by @code{VALUE} used by @code{TO}:
                   2104: @code{-32 throw} (Invalid name argument)
                   2105: 
1.15      anton    2106: @item name not found (@code{'}, @code{POSTPONE}, @code{[']}, @code{[COMPILE]}):
1.14      anton    2107: @code{-13 throw} (Undefined word)
                   2108: 
                   2109: @item parameters are not of the same type (@code{DO}, @code{?DO}, @code{WITHIN}):
                   2110: Gforth behaves as if they were of the same type. I.e., you can predict
                   2111: the behaviour by interpreting all parameters as, e.g., signed.
                   2112: 
                   2113: @item @code{POSTPONE} or @code{[COMPILE]} applied to @code{TO}:
                   2114: Assume @code{: X POSTPONE TO ; IMMEDIATE}. @code{X} is equivalent to
                   2115: @code{TO}.
                   2116: 
                   2117: @item String longer than a counted string returned by @code{WORD}:
                   2118: Not checked. The string will be ok, but the count will, of course,
                   2119: contain only the least significant bits of the length.
                   2120: 
1.15      anton    2121: @item u greater than or equal to the number of bits in a cell (@code{LSHIFT}, @code{RSHIFT}):
1.14      anton    2122: Processor-dependent. Typical behaviours are returning 0 and using only
                   2123: the low bits of the shift count.
                   2124: 
                   2125: @item word not defined via @code{CREATE}:
                   2126: @code{>BODY} produces the PFA of the word no matter how it was defined.
                   2127: 
                   2128: @code{DOES>} changes the execution semantics of the last defined word no
                   2129: matter how it was defined. E.g., @code{CONSTANT DOES>} is equivalent to
                   2130: @code{CREATE , DOES>}.
                   2131: 
                   2132: @item words improperly used outside @code{<#} and @code{#>}:
                   2133: Not checked. As usual, you can expect memory faults.
                   2134: 
                   2135: @end table
                   2136: 
                   2137: 
                   2138: @c ---------------------------------------------------------------------
                   2139: @node core-other,  , core-ambcond, The Core Words
                   2140: @subsection Other system documentation
                   2141: @c ---------------------------------------------------------------------
                   2142: 
                   2143: @table @i
                   2144: 
                   2145: @item nonstandard words using @code{PAD}:
                   2146: None.
                   2147: 
                   2148: @item operator's terminal facilities available:
                   2149: !!??
                   2150: 
                   2151: @item program data space available:
                   2152: @code{sp@ here - .} gives the space remaining for dictionary and data
                   2153: stack together.
                   2154: 
                   2155: @item return stack space available:
                   2156: !!??
                   2157: 
                   2158: @item stack space available:
                   2159: @code{sp@ here - .} gives the space remaining for dictionary and data
                   2160: stack together.
                   2161: 
                   2162: @item system dictionary space required, in address units:
                   2163: Type @code{here forthstart - .} after startup. At the time of this
                   2164: writing, this gives 70108 (bytes) on a 32-bit system.
                   2165: @end table
                   2166: 
                   2167: 
                   2168: @c =====================================================================
                   2169: @node The optional Block word set, The optional Double Number word set, The Core Words, ANS conformance
                   2170: @section The optional Block word set
                   2171: @c =====================================================================
                   2172: 
                   2173: @menu
1.15      anton    2174: * block-idef::                  Implementation Defined Options                  
                   2175: * block-ambcond::               Ambiguous Conditions               
                   2176: * block-other::                 Other System Documentation                 
1.14      anton    2177: @end menu
                   2178: 
                   2179: 
                   2180: @c ---------------------------------------------------------------------
                   2181: @node block-idef, block-ambcond, The optional Block word set, The optional Block word set
                   2182: @subsection Implementation Defined Options
                   2183: @c ---------------------------------------------------------------------
                   2184: 
                   2185: @table @i
                   2186: 
                   2187: @item the format for display by @code{LIST}:
                   2188: First the screen number is displayed, then 16 lines of 64 characters,
                   2189: each line preceded by the line number.
                   2190: 
                   2191: @item the length of a line affected by @code{\}:
                   2192: 64 characters.
                   2193: @end table
                   2194: 
                   2195: 
                   2196: @c ---------------------------------------------------------------------
                   2197: @node block-ambcond, block-other, block-idef, The optional Block word set
                   2198: @subsection Ambiguous conditions
                   2199: @c ---------------------------------------------------------------------
                   2200: 
                   2201: @table @i
                   2202: 
                   2203: @item correct block read was not possible:
                   2204: Typically results in a @code{throw} of some OS-derived value (between
                   2205: -512 and -2048). If the blocks file was just not long enough, blanks are
                   2206: supplied for the missing portion.
                   2207: 
                   2208: @item I/O exception in block transfer:
                   2209: Typically results in a @code{throw} of some OS-derived value (between
                   2210: -512 and -2048).
                   2211: 
                   2212: @item invalid block number:
                   2213: @code{-35 throw} (Invalid block number)
                   2214: 
                   2215: @item a program directly alters the contents of @code{BLK}:
                   2216: The input stream is switched to that other block, at the same
                   2217: position. If the storing to @code{BLK} happens when interpreting
                   2218: non-block input, the system will get quite confused when the block ends.
                   2219: 
                   2220: @item no current block buffer for @code{UPDATE}:
                   2221: @code{UPDATE} has no effect.
                   2222: 
                   2223: @end table
                   2224: 
                   2225: 
                   2226: @c ---------------------------------------------------------------------
                   2227: @node block-other,  , block-ambcond, The optional Block word set
                   2228: @subsection Other system documentation
                   2229: @c ---------------------------------------------------------------------
                   2230: 
                   2231: @table @i
                   2232: 
                   2233: @item any restrictions a multiprogramming system places on the use of buffer addresses:
                   2234: No restrictions (yet).
                   2235: 
                   2236: @item the number of blocks available for source and data:
                   2237: depends on your disk space.
                   2238: 
                   2239: @end table
                   2240: 
                   2241: 
                   2242: @c =====================================================================
                   2243: @node The optional Double Number word set, The optional Exception word set, The optional Block word set, ANS conformance
                   2244: @section The optional Double Number word set
                   2245: @c =====================================================================
                   2246: 
                   2247: @menu
1.15      anton    2248: * double-ambcond::              Ambiguous Conditions              
1.14      anton    2249: @end menu
                   2250: 
                   2251: 
                   2252: @c ---------------------------------------------------------------------
1.15      anton    2253: @node double-ambcond,  , The optional Double Number word set, The optional Double Number word set
1.14      anton    2254: @subsection Ambiguous conditions
                   2255: @c ---------------------------------------------------------------------
                   2256: 
                   2257: @table @i
                   2258: 
1.15      anton    2259: @item @var{d} outside of range of @var{n} in @code{D>S}:
1.14      anton    2260: The least significant cell of @var{d} is produced.
                   2261: 
                   2262: @end table
                   2263: 
                   2264: 
                   2265: @c =====================================================================
                   2266: @node The optional Exception word set, The optional Facility word set, The optional Double Number word set, ANS conformance
                   2267: @section The optional Exception word set
                   2268: @c =====================================================================
                   2269: 
                   2270: @menu
1.15      anton    2271: * exception-idef::              Implementation Defined Options              
1.14      anton    2272: @end menu
                   2273: 
                   2274: 
                   2275: @c ---------------------------------------------------------------------
1.15      anton    2276: @node exception-idef,  , The optional Exception word set, The optional Exception word set
1.14      anton    2277: @subsection Implementation Defined Options
                   2278: @c ---------------------------------------------------------------------
                   2279: 
                   2280: @table @i
                   2281: @item @code{THROW}-codes used in the system:
                   2282: The codes -256@minus{}-511 are used for reporting signals (see
                   2283: @file{errore.fs}). The codes -512@minus{}-2047 are used for OS errors
                   2284: (for file and memory allocation operations). The mapping from OS error
                   2285: numbers to throw code is -512@minus{}@var{errno}. One side effect of
                   2286: this mapping is that undefined OS errors produce a message with a
                   2287: strange number; e.g., @code{-1000 THROW} results in @code{Unknown error
                   2288: 488} on my system.
                   2289: @end table
                   2290: 
                   2291: @c =====================================================================
                   2292: @node The optional Facility word set, The optional File-Access word set, The optional Exception word set, ANS conformance
                   2293: @section The optional Facility word set
                   2294: @c =====================================================================
                   2295: 
                   2296: @menu
1.15      anton    2297: * facility-idef::               Implementation Defined Options               
                   2298: * facility-ambcond::            Ambiguous Conditions            
1.14      anton    2299: @end menu
                   2300: 
                   2301: 
                   2302: @c ---------------------------------------------------------------------
                   2303: @node facility-idef, facility-ambcond, The optional Facility word set, The optional Facility word set
                   2304: @subsection Implementation Defined Options
                   2305: @c ---------------------------------------------------------------------
                   2306: 
                   2307: @table @i
                   2308: 
                   2309: @item encoding of keyboard events (@code{EKEY}):
                   2310: Not yet implemeted.
                   2311: 
                   2312: @item duration of a system clock tick
                   2313: System dependent. With respect to @code{MS}, the time is specified in
                   2314: microseconds. How well the OS and the hardware implement this, is
                   2315: another question.
                   2316: 
                   2317: @item repeatability to be expected from the execution of @code{MS}:
                   2318: System dependent. On Unix, a lot depends on load. If the system is
1.17      anton    2319: lightly loaded, and the delay is short enough that Gforth does not get
1.14      anton    2320: swapped out, the performance should be acceptable. Under MS-DOS and
                   2321: other single-tasking systems, it should be good.
                   2322: 
                   2323: @end table
                   2324: 
                   2325: 
                   2326: @c ---------------------------------------------------------------------
1.15      anton    2327: @node facility-ambcond,  , facility-idef, The optional Facility word set
1.14      anton    2328: @subsection Ambiguous conditions
                   2329: @c ---------------------------------------------------------------------
                   2330: 
                   2331: @table @i
                   2332: 
                   2333: @item @code{AT-XY} can't be performed on user output device:
                   2334: Largely terminal dependant. No range checks are done on the arguments.
                   2335: No errors are reported. You may see some garbage appearing, you may see
                   2336: simply nothing happen.
                   2337: 
                   2338: @end table
                   2339: 
                   2340: 
                   2341: @c =====================================================================
                   2342: @node The optional File-Access word set, The optional Floating-Point word set, The optional Facility word set, ANS conformance
                   2343: @section The optional File-Access word set
                   2344: @c =====================================================================
                   2345: 
                   2346: @menu
1.15      anton    2347: * file-idef::                   Implementation Defined Options                   
                   2348: * file-ambcond::                Ambiguous Conditions                
1.14      anton    2349: @end menu
                   2350: 
                   2351: 
                   2352: @c ---------------------------------------------------------------------
                   2353: @node file-idef, file-ambcond, The optional File-Access word set, The optional File-Access word set
                   2354: @subsection Implementation Defined Options
                   2355: @c ---------------------------------------------------------------------
                   2356: 
                   2357: @table @i
                   2358: 
                   2359: @item File access methods used:
                   2360: @code{R/O}, @code{R/W} and @code{BIN} work as you would
                   2361: expect. @code{W/O} translates into the C file opening mode @code{w} (or
                   2362: @code{wb}): The file is cleared, if it exists, and created, if it does
1.15      anton    2363: not (both with @code{open-file} and @code{create-file}).  Under Unix
1.14      anton    2364: @code{create-file} creates a file with 666 permissions modified by your
                   2365: umask.
                   2366: 
                   2367: @item file exceptions:
                   2368: The file words do not raise exceptions (except, perhaps, memory access
                   2369: faults when you pass illegal addresses or file-ids).
                   2370: 
                   2371: @item file line terminator:
                   2372: System-dependent. Gforth uses C's newline character as line
                   2373: terminator. What the actual character code(s) of this are is
                   2374: system-dependent.
                   2375: 
                   2376: @item file name format
                   2377: System dependent. Gforth just uses the file name format of your OS.
                   2378: 
                   2379: @item information returned by @code{FILE-STATUS}:
                   2380: @code{FILE-STATUS} returns the most powerful file access mode allowed
                   2381: for the file: Either @code{R/O}, @code{W/O} or @code{R/W}. If the file
                   2382: cannot be accessed, @code{R/O BIN} is returned. @code{BIN} is applicable
                   2383: along with the retured mode.
                   2384: 
                   2385: @item input file state after an exception when including source:
                   2386: All files that are left via the exception are closed.
                   2387: 
                   2388: @item @var{ior} values and meaning:
1.15      anton    2389: The @var{ior}s returned by the file and memory allocation words are
                   2390: intended as throw codes. They typically are in the range
                   2391: -512@minus{}-2047 of OS errors.  The mapping from OS error numbers to
                   2392: @var{ior}s is -512@minus{}@var{errno}.
1.14      anton    2393: 
                   2394: @item maximum depth of file input nesting:
                   2395: limited by the amount of return stack, locals/TIB stack, and the number
                   2396: of open files available. This should not give you troubles.
                   2397: 
                   2398: @item maximum size of input line:
                   2399: @code{/line}. Currently 255.
                   2400: 
                   2401: @item methods of mapping block ranges to files:
                   2402: Currently, the block words automatically access the file
                   2403: @file{blocks.fb} in the currend working directory. More sophisticated
                   2404: methods could be implemented if there is demand (and a volunteer).
                   2405: 
                   2406: @item number of string buffers provided by @code{S"}:
                   2407: 1
                   2408: 
                   2409: @item size of string buffer used by @code{S"}:
                   2410: @code{/line}. currently 255.
                   2411: 
                   2412: @end table
                   2413: 
                   2414: @c ---------------------------------------------------------------------
1.15      anton    2415: @node file-ambcond,  , file-idef, The optional File-Access word set
1.14      anton    2416: @subsection Ambiguous conditions
                   2417: @c ---------------------------------------------------------------------
                   2418: 
                   2419: @table @i
                   2420: 
                   2421: @item attempting to position a file outside it's boundaries:
                   2422: @code{REPOSITION-FILE} is performed as usual: Afterwards,
                   2423: @code{FILE-POSITION} returns the value given to @code{REPOSITION-FILE}.
                   2424: 
                   2425: @item attempting to read from file positions not yet written:
                   2426: End-of-file, i.e., zero characters are read and no error is reported.
                   2427: 
                   2428: @item @var{file-id} is invalid (@code{INCLUDE-FILE}):
                   2429: An appropriate exception may be thrown, but a memory fault or other
                   2430: problem is more probable.
                   2431: 
                   2432: @item I/O exception reading or closing @var{file-id} (@code{include-file}, @code{included}):
                   2433: The @var{ior} produced by the operation, that discovered the problem, is
                   2434: thrown.
                   2435: 
                   2436: @item named file cannot be opened (@code{included}):
                   2437: The @var{ior} produced by @code{open-file} is thrown.
                   2438: 
                   2439: @item requesting an unmapped block number:
                   2440: There are no unmapped legal block numbers. On some operating systems,
                   2441: writing a block with a large number may overflow the file system and
                   2442: have an error message as consequence.
                   2443: 
                   2444: @item using @code{source-id} when @code{blk} is non-zero:
                   2445: @code{source-id} performs its function. Typically it will give the id of
                   2446: the source which loaded the block. (Better ideas?)
                   2447: 
                   2448: @end table
                   2449: 
                   2450: 
                   2451: @c =====================================================================
                   2452: @node  The optional Floating-Point word set, The optional Locals word set, The optional File-Access word set, ANS conformance
1.15      anton    2453: @section The optional Floating-Point word set
1.14      anton    2454: @c =====================================================================
                   2455: 
                   2456: @menu
1.15      anton    2457: * floating-idef::               Implementation Defined Options
                   2458: * floating-ambcond::            Ambiguous Conditions            
1.14      anton    2459: @end menu
                   2460: 
                   2461: 
                   2462: @c ---------------------------------------------------------------------
                   2463: @node floating-idef, floating-ambcond, The optional Floating-Point word set, The optional Floating-Point word set
                   2464: @subsection Implementation Defined Options
                   2465: @c ---------------------------------------------------------------------
                   2466: 
                   2467: @table @i
                   2468: 
1.15      anton    2469: @item format and range of floating point numbers:
                   2470: System-dependent; the @code{double} type of C.
1.14      anton    2471: 
1.15      anton    2472: @item results of @code{REPRESENT} when @var{float} is out of range:
                   2473: System dependent; @code{REPRESENT} is implemented using the C library
                   2474: function @code{ecvt()} and inherits its behaviour in this respect.
1.14      anton    2475: 
1.15      anton    2476: @item rounding or truncation of floating-point numbers:
                   2477: What's the question?!!
1.14      anton    2478: 
1.15      anton    2479: @item size of floating-point stack:
                   2480: @code{s" FLOATING-STACK" environment? drop .}. Can be changed at startup
                   2481: with the command-line option @code{-f}.
1.14      anton    2482: 
1.15      anton    2483: @item width of floating-point stack:
                   2484: @code{1 floats}.
1.14      anton    2485: 
                   2486: @end table
                   2487: 
                   2488: 
                   2489: @c ---------------------------------------------------------------------
1.15      anton    2490: @node floating-ambcond,  , floating-idef, The optional Floating-Point word set
                   2491: @subsection Ambiguous conditions
1.14      anton    2492: @c ---------------------------------------------------------------------
                   2493: 
                   2494: @table @i
                   2495: 
1.15      anton    2496: @item @code{df@@} or @code{df!} used with an address that is not double-float  aligned:
                   2497: System-dependent. Typically results in an alignment fault like other
                   2498: alignment violations.
1.14      anton    2499: 
1.15      anton    2500: @item @code{f@@} or @code{f!} used with an address that is not float  aligned:
                   2501: System-dependent. Typically results in an alignment fault like other
                   2502: alignment violations.
1.14      anton    2503: 
1.15      anton    2504: @item Floating-point result out of range:
                   2505: System-dependent. Can result in a @code{-55 THROW} (Floating-point
                   2506: unidentified fault), or can produce a special value representing, e.g.,
                   2507: Infinity.
1.14      anton    2508: 
1.15      anton    2509: @item @code{sf@@} or @code{sf!} used with an address that is not single-float  aligned:
                   2510: System-dependent. Typically results in an alignment fault like other
                   2511: alignment violations.
1.14      anton    2512: 
1.15      anton    2513: @item BASE is not decimal (@code{REPRESENT}, @code{F.}, @code{FE.}, @code{FS.}):
                   2514: The floating-point number is converted into decimal nonetheless.
1.14      anton    2515: 
1.15      anton    2516: @item Both arguments are equal to zero (@code{FATAN2}):
                   2517: System-dependent. @code{FATAN2} is implemented using the C library
                   2518: function @code{atan2()}.
1.14      anton    2519: 
1.15      anton    2520: @item Using ftan on an argument @var{r1} where cos(@var{r1}) is zero:
                   2521: System-dependent. Anyway, typically the cos of @var{r1} will not be zero
                   2522: because of small errors and the tan will be a very large (or very small)
                   2523: but finite number.
1.14      anton    2524: 
1.15      anton    2525: @item @var{d} cannot be presented precisely as a float in @code{D>F}:
                   2526: The result is rounded to the nearest float.
1.14      anton    2527: 
1.15      anton    2528: @item dividing by zero:
                   2529: @code{-55 throw} (Floating-point unidentified fault)
1.14      anton    2530: 
1.15      anton    2531: @item exponent too big for conversion (@code{DF!}, @code{DF@@}, @code{SF!}, @code{SF@@}):
                   2532: System dependent. On IEEE-FP based systems the number is converted into
                   2533: an infinity.
1.14      anton    2534: 
1.15      anton    2535: @item @var{float}<1 (@code{facosh}):
                   2536: @code{-55 throw} (Floating-point unidentified fault)
1.14      anton    2537: 
1.15      anton    2538: @item @var{float}=<-1 (@code{flnp1}):
                   2539: @code{-55 throw} (Floating-point unidentified fault). On IEEE-FP systems
                   2540: negative infinity is typically produced for @var{float}=-1.
1.14      anton    2541: 
1.15      anton    2542: @item @var{float}=<0 (@code{fln}, @code{flog}):
                   2543: @code{-55 throw} (Floating-point unidentified fault). On IEEE-FP systems
                   2544: negative infinity is typically produced for @var{float}=0.
1.14      anton    2545: 
1.15      anton    2546: @item @var{float}<0 (@code{fasinh}, @code{fsqrt}):
                   2547: @code{-55 throw} (Floating-point unidentified fault). @code{fasinh}
                   2548: produces values for these inputs on my Linux box (Bug in the C library?)
1.14      anton    2549: 
1.15      anton    2550: @item |@var{float}|>1 (@code{facos}, @code{fasin}, @code{fatanh}):
                   2551: @code{-55 throw} (Floating-point unidentified fault).
1.14      anton    2552: 
1.15      anton    2553: @item integer part of float cannot be represented by @var{d} in @code{f>d}:
                   2554: @code{-55 throw} (Floating-point unidentified fault).
1.14      anton    2555: 
1.15      anton    2556: @item string larger than pictured numeric output area (@code{f.}, @code{fe.}, @code{fs.}):
                   2557: This does not happen.
                   2558: @end table
1.14      anton    2559: 
                   2560: 
                   2561: 
                   2562: @c =====================================================================
1.15      anton    2563: @node  The optional Locals word set, The optional Memory-Allocation word set, The optional Floating-Point word set, ANS conformance
                   2564: @section The optional Locals word set
1.14      anton    2565: @c =====================================================================
                   2566: 
                   2567: @menu
1.15      anton    2568: * locals-idef::                 Implementation Defined Options                 
                   2569: * locals-ambcond::              Ambiguous Conditions              
1.14      anton    2570: @end menu
                   2571: 
                   2572: 
                   2573: @c ---------------------------------------------------------------------
1.15      anton    2574: @node locals-idef, locals-ambcond, The optional Locals word set, The optional Locals word set
1.14      anton    2575: @subsection Implementation Defined Options
                   2576: @c ---------------------------------------------------------------------
                   2577: 
                   2578: @table @i
                   2579: 
1.15      anton    2580: @item maximum number of locals in a definition:
                   2581: @code{s" #locals" environment? drop .}. Currently 15. This is a lower
                   2582: bound, e.g., on a 32-bit machine there can be 41 locals of up to 8
                   2583: characters. The number of locals in a definition is bounded by the size
                   2584: of locals-buffer, which contains the names of the locals.
1.14      anton    2585: 
                   2586: @end table
                   2587: 
                   2588: 
                   2589: @c ---------------------------------------------------------------------
1.15      anton    2590: @node locals-ambcond,  , locals-idef, The optional Locals word set
1.14      anton    2591: @subsection Ambiguous conditions
                   2592: @c ---------------------------------------------------------------------
                   2593: 
                   2594: @table @i
                   2595: 
1.15      anton    2596: @item executing a named local in interpretation state:
                   2597: @code{-14 throw} (Interpreting a compile-only word).
1.14      anton    2598: 
1.15      anton    2599: @item @var{name} not defined by @code{VALUE} or @code{(LOCAL)} (@code{TO}):
                   2600: @code{-32 throw} (Invalid name argument)
1.14      anton    2601: 
                   2602: @end table
                   2603: 
                   2604: 
                   2605: @c =====================================================================
1.15      anton    2606: @node  The optional Memory-Allocation word set, The optional Programming-Tools word set, The optional Locals word set, ANS conformance
                   2607: @section The optional Memory-Allocation word set
1.14      anton    2608: @c =====================================================================
                   2609: 
                   2610: @menu
1.15      anton    2611: * memory-idef::                 Implementation Defined Options                 
1.14      anton    2612: @end menu
                   2613: 
                   2614: 
                   2615: @c ---------------------------------------------------------------------
1.15      anton    2616: @node memory-idef,  , The optional Memory-Allocation word set, The optional Memory-Allocation word set
1.14      anton    2617: @subsection Implementation Defined Options
                   2618: @c ---------------------------------------------------------------------
                   2619: 
                   2620: @table @i
                   2621: 
1.15      anton    2622: @item values and meaning of @var{ior}:
                   2623: The @var{ior}s returned by the file and memory allocation words are
                   2624: intended as throw codes. They typically are in the range
                   2625: -512@minus{}-2047 of OS errors.  The mapping from OS error numbers to
                   2626: @var{ior}s is -512@minus{}@var{errno}.
1.14      anton    2627: 
                   2628: @end table
                   2629: 
                   2630: @c =====================================================================
1.15      anton    2631: @node  The optional Programming-Tools word set, The optional Search-Order word set, The optional Memory-Allocation word set, ANS conformance
                   2632: @section The optional Programming-Tools word set
1.14      anton    2633: @c =====================================================================
                   2634: 
                   2635: @menu
1.15      anton    2636: * programming-idef::            Implementation Defined Options            
                   2637: * programming-ambcond::         Ambiguous Conditions         
1.14      anton    2638: @end menu
                   2639: 
                   2640: 
                   2641: @c ---------------------------------------------------------------------
1.15      anton    2642: @node programming-idef, programming-ambcond, The optional Programming-Tools word set, The optional Programming-Tools word set
1.14      anton    2643: @subsection Implementation Defined Options
                   2644: @c ---------------------------------------------------------------------
                   2645: 
                   2646: @table @i
                   2647: 
1.15      anton    2648: @item ending sequence for input following @code{;code} and @code{code}:
                   2649: Not implemented (yet).
1.14      anton    2650: 
1.15      anton    2651: @item manner of processing input following @code{;code} and @code{code}:
                   2652: Not implemented (yet).
                   2653: 
                   2654: @item search order capability for @code{EDITOR} and @code{ASSEMBLER}:
                   2655: Not implemented (yet). If they were implemented, they would use the
                   2656: search order wordset.
                   2657: 
                   2658: @item source and format of display by @code{SEE}:
                   2659: The source for @code{see} is the intermediate code used by the inner
                   2660: interpreter.  The current @code{see} tries to output Forth source code
                   2661: as well as possible.
                   2662: 
1.14      anton    2663: @end table
                   2664: 
                   2665: @c ---------------------------------------------------------------------
1.15      anton    2666: @node programming-ambcond,  , programming-idef, The optional Programming-Tools word set
1.14      anton    2667: @subsection Ambiguous conditions
                   2668: @c ---------------------------------------------------------------------
                   2669: 
                   2670: @table @i
                   2671: 
1.15      anton    2672: @item deleting the compilation wordlist (@code{FORGET}):
                   2673: Not implemented (yet).
1.14      anton    2674: 
1.15      anton    2675: @item fewer than @var{u}+1 items on the control flow stack (@code{CS-PICK}, @code{CS-ROLL}):
                   2676: This typically results in an @code{abort"} with a descriptive error
                   2677: message (may change into a @code{-22 throw} (Control structure mismatch)
                   2678: in the future). You may also get a memory access error. If you are
                   2679: unlucky, this ambiguous condition is not caught.
                   2680: 
                   2681: @item @var{name} can't be found (@code{forget}):
                   2682: Not implemented (yet).
1.14      anton    2683: 
1.15      anton    2684: @item @var{name} not defined via @code{CREATE}:
                   2685: @code{;code} is not implemented (yet). If it were, it would behave like
                   2686: @code{DOES>} in this respect, i.e., change the execution semantics of
                   2687: the last defined word no matter how it was defined.
1.14      anton    2688: 
1.15      anton    2689: @item @code{POSTPONE} applied to @code{[IF]}:
                   2690: After defining @code{: X POSTPONE [IF] ; IMMEDIATE}. @code{X} is
                   2691: equivalent to @code{[IF]}.
1.14      anton    2692: 
1.15      anton    2693: @item reaching the end of the input source before matching @code{[ELSE]} or @code{[THEN]}:
                   2694: Continue in the same state of conditional compilation in the next outer
                   2695: input source. Currently there is no warning to the user about this.
1.14      anton    2696: 
1.15      anton    2697: @item removing a needed definition (@code{FORGET}):
                   2698: Not implemented (yet).
1.14      anton    2699: 
                   2700: @end table
                   2701: 
                   2702: 
                   2703: @c =====================================================================
1.15      anton    2704: @node  The optional Search-Order word set,  , The optional Programming-Tools word set, ANS conformance
                   2705: @section The optional Search-Order word set
1.14      anton    2706: @c =====================================================================
                   2707: 
                   2708: @menu
1.15      anton    2709: * search-idef::                 Implementation Defined Options                 
                   2710: * search-ambcond::              Ambiguous Conditions              
1.14      anton    2711: @end menu
                   2712: 
                   2713: 
                   2714: @c ---------------------------------------------------------------------
1.15      anton    2715: @node search-idef, search-ambcond, The optional Search-Order word set, The optional Search-Order word set
1.14      anton    2716: @subsection Implementation Defined Options
                   2717: @c ---------------------------------------------------------------------
                   2718: 
                   2719: @table @i
                   2720: 
1.15      anton    2721: @item maximum number of word lists in search order:
                   2722: @code{s" wordlists" environment? drop .}. Currently 16.
                   2723: 
                   2724: @item minimum search order:
                   2725: @code{root root}.
1.14      anton    2726: 
                   2727: @end table
                   2728: 
                   2729: @c ---------------------------------------------------------------------
1.15      anton    2730: @node search-ambcond,  , search-idef, The optional Search-Order word set
1.14      anton    2731: @subsection Ambiguous conditions
                   2732: @c ---------------------------------------------------------------------
                   2733: 
                   2734: @table @i
                   2735: 
1.15      anton    2736: @item changing the compilation wordlist (during compilation):
                   2737: The definition is put into the wordlist that is the compilation wordlist
                   2738: when @code{REVEAL} is executed (by @code{;}, @code{DOES>},
                   2739: @code{RECURSIVE}, etc.).
1.14      anton    2740: 
1.15      anton    2741: @item search order empty (@code{previous}):
                   2742: @code{abort" Vocstack empty"}.
1.14      anton    2743: 
1.15      anton    2744: @item too many word lists in search order (@code{also}):
                   2745: @code{abort" Vocstack full"}.
1.14      anton    2746: 
                   2747: @end table
1.13      anton    2748: 
                   2749: 
1.17      anton    2750: @node Model, Emacs and Gforth, ANS conformance, Top
1.4       anton    2751: @chapter Model
                   2752: 
1.17      anton    2753: @node Emacs and Gforth, Internals, Model, Top
                   2754: @chapter Emacs and Gforth
1.4       anton    2755: 
1.17      anton    2756: Gforth comes with @file{gforth.el}, an improved version of
1.4       anton    2757: @file{forth.el} by Goran Rydqvist (icluded in the TILE package). The
                   2758: improvements are a better (but still not perfect) handling of
                   2759: indentation. I have also added comment paragraph filling (@kbd{M-q}),
1.8       anton    2760: commenting (@kbd{C-x \}) and uncommenting (@kbd{C-u C-x \}) regions and
                   2761: removing debugging tracers (@kbd{C-x ~}, @pxref{Debugging}). I left the
                   2762: stuff I do not use alone, even though some of it only makes sense for
                   2763: TILE. To get a description of these features, enter Forth mode and type
                   2764: @kbd{C-h m}.
1.4       anton    2765: 
1.17      anton    2766: In addition, Gforth supports Emacs quite well: The source code locations
1.4       anton    2767: given in error messages, debugging output (from @code{~~}) and failed
                   2768: assertion messages are in the right format for Emacs' compilation mode
                   2769: (@pxref{Compilation, , Running Compilations under Emacs, emacs, Emacs
                   2770: Manual}) so the source location corresponding to an error or other
                   2771: message is only a few keystrokes away (@kbd{C-x `} for the next error,
                   2772: @kbd{C-c C-c} for the error under the cursor).
                   2773: 
                   2774: Also, if you @code{include} @file{etags.fs}, a new @file{TAGS} file
                   2775: (@pxref{Tags, , Tags Tables, emacs, Emacs Manual}) will be produced that
                   2776: contains the definitions of all words defined afterwards. You can then
                   2777: find the source for a word using @kbd{M-.}. Note that emacs can use
1.17      anton    2778: several tags files at the same time (e.g., one for the Gforth sources
1.4       anton    2779: and one for your program).
                   2780: 
                   2781: To get all these benefits, add the following lines to your @file{.emacs}
                   2782: file:
                   2783: 
                   2784: @example
                   2785: (autoload 'forth-mode "gforth.el")
                   2786: (setq auto-mode-alist (cons '("\\.fs\\'" . forth-mode) auto-mode-alist))
                   2787: @end example
                   2788: 
1.17      anton    2789: @node Internals, Bugs, Emacs and Gforth, Top
1.3       anton    2790: @chapter Internals
                   2791: 
1.17      anton    2792: Reading this section is not necessary for programming with Gforth. It
                   2793: should be helpful for finding your way in the Gforth sources.
1.3       anton    2794: 
1.4       anton    2795: @menu
                   2796: * Portability::                 
                   2797: * Threading::                   
                   2798: * Primitives::                  
                   2799: * System Architecture::         
1.17      anton    2800: * Performance::                 
1.4       anton    2801: @end menu
                   2802: 
                   2803: @node Portability, Threading, Internals, Internals
1.3       anton    2804: @section Portability
                   2805: 
                   2806: One of the main goals of the effort is availability across a wide range
                   2807: of personal machines. fig-Forth, and, to a lesser extent, F83, achieved
                   2808: this goal by manually coding the engine in assembly language for several
                   2809: then-popular processors. This approach is very labor-intensive and the
                   2810: results are short-lived due to progress in computer architecture.
                   2811: 
                   2812: Others have avoided this problem by coding in C, e.g., Mitch Bradley
                   2813: (cforth), Mikael Patel (TILE) and Dirk Zoller (pfe). This approach is
                   2814: particularly popular for UNIX-based Forths due to the large variety of
                   2815: architectures of UNIX machines. Unfortunately an implementation in C
                   2816: does not mix well with the goals of efficiency and with using
                   2817: traditional techniques: Indirect or direct threading cannot be expressed
                   2818: in C, and switch threading, the fastest technique available in C, is
                   2819: significantly slower. Another problem with C is that it's very
                   2820: cumbersome to express double integer arithmetic.
                   2821: 
                   2822: Fortunately, there is a portable language that does not have these
                   2823: limitations: GNU C, the version of C processed by the GNU C compiler
                   2824: (@pxref{C Extensions, , Extensions to the C Language Family, gcc.info,
                   2825: GNU C Manual}). Its labels as values feature (@pxref{Labels as Values, ,
                   2826: Labels as Values, gcc.info, GNU C Manual}) makes direct and indirect
                   2827: threading possible, its @code{long long} type (@pxref{Long Long, ,
                   2828: Double-Word Integers, gcc.info, GNU C Manual}) corresponds to Forths
                   2829: double numbers. GNU C is available for free on all important (and many
                   2830: unimportant) UNIX machines, VMS, 80386s running MS-DOS, the Amiga, and
                   2831: the Atari ST, so a Forth written in GNU C can run on all these
1.17      anton    2832: machines.
1.3       anton    2833: 
                   2834: Writing in a portable language has the reputation of producing code that
                   2835: is slower than assembly. For our Forth engine we repeatedly looked at
                   2836: the code produced by the compiler and eliminated most compiler-induced
                   2837: inefficiencies by appropriate changes in the source-code.
                   2838: 
                   2839: However, register allocation cannot be portably influenced by the
                   2840: programmer, leading to some inefficiencies on register-starved
                   2841: machines. We use explicit register declarations (@pxref{Explicit Reg
                   2842: Vars, , Variables in Specified Registers, gcc.info, GNU C Manual}) to
                   2843: improve the speed on some machines. They are turned on by using the
                   2844: @code{gcc} switch @code{-DFORCE_REG}. Unfortunately, this feature not
                   2845: only depends on the machine, but also on the compiler version: On some
                   2846: machines some compiler versions produce incorrect code when certain
                   2847: explicit register declarations are used. So by default
                   2848: @code{-DFORCE_REG} is not used.
                   2849: 
1.4       anton    2850: @node Threading, Primitives, Portability, Internals
1.3       anton    2851: @section Threading
                   2852: 
                   2853: GNU C's labels as values extension (available since @code{gcc-2.0},
                   2854: @pxref{Labels as Values, , Labels as Values, gcc.info, GNU C Manual})
                   2855: makes it possible to take the address of @var{label} by writing
                   2856: @code{&&@var{label}}.  This address can then be used in a statement like
                   2857: @code{goto *@var{address}}. I.e., @code{goto *&&x} is the same as
                   2858: @code{goto x}.
                   2859: 
                   2860: With this feature an indirect threaded NEXT looks like:
                   2861: @example
                   2862: cfa = *ip++;
                   2863: ca = *cfa;
                   2864: goto *ca;
                   2865: @end example
                   2866: For those unfamiliar with the names: @code{ip} is the Forth instruction
                   2867: pointer; the @code{cfa} (code-field address) corresponds to ANS Forths
                   2868: execution token and points to the code field of the next word to be
                   2869: executed; The @code{ca} (code address) fetched from there points to some
                   2870: executable code, e.g., a primitive or the colon definition handler
                   2871: @code{docol}.
                   2872: 
                   2873: Direct threading is even simpler:
                   2874: @example
                   2875: ca = *ip++;
                   2876: goto *ca;
                   2877: @end example
                   2878: 
                   2879: Of course we have packaged the whole thing neatly in macros called
                   2880: @code{NEXT} and @code{NEXT1} (the part of NEXT after fetching the cfa).
                   2881: 
1.4       anton    2882: @menu
                   2883: * Scheduling::                  
                   2884: * Direct or Indirect Threaded?::  
                   2885: * DOES>::                       
                   2886: @end menu
                   2887: 
                   2888: @node Scheduling, Direct or Indirect Threaded?, Threading, Threading
1.3       anton    2889: @subsection Scheduling
                   2890: 
                   2891: There is a little complication: Pipelined and superscalar processors,
                   2892: i.e., RISC and some modern CISC machines can process independent
                   2893: instructions while waiting for the results of an instruction. The
                   2894: compiler usually reorders (schedules) the instructions in a way that
                   2895: achieves good usage of these delay slots. However, on our first tries
                   2896: the compiler did not do well on scheduling primitives. E.g., for
                   2897: @code{+} implemented as
                   2898: @example
                   2899: n=sp[0]+sp[1];
                   2900: sp++;
                   2901: sp[0]=n;
                   2902: NEXT;
                   2903: @end example
                   2904: the NEXT comes strictly after the other code, i.e., there is nearly no
                   2905: scheduling. After a little thought the problem becomes clear: The
                   2906: compiler cannot know that sp and ip point to different addresses (and
1.4       anton    2907: the version of @code{gcc} we used would not know it even if it was
                   2908: possible), so it could not move the load of the cfa above the store to
                   2909: the TOS. Indeed the pointers could be the same, if code on or very near
                   2910: the top of stack were executed. In the interest of speed we chose to
                   2911: forbid this probably unused ``feature'' and helped the compiler in
                   2912: scheduling: NEXT is divided into the loading part (@code{NEXT_P1}) and
                   2913: the goto part (@code{NEXT_P2}). @code{+} now looks like:
1.3       anton    2914: @example
                   2915: n=sp[0]+sp[1];
                   2916: sp++;
                   2917: NEXT_P1;
                   2918: sp[0]=n;
                   2919: NEXT_P2;
                   2920: @end example
1.4       anton    2921: This can be scheduled optimally by the compiler.
1.3       anton    2922: 
                   2923: This division can be turned off with the switch @code{-DCISC_NEXT}. This
                   2924: switch is on by default on machines that do not profit from scheduling
                   2925: (e.g., the 80386), in order to preserve registers.
                   2926: 
1.4       anton    2927: @node Direct or Indirect Threaded?, DOES>, Scheduling, Threading
1.3       anton    2928: @subsection Direct or Indirect Threaded?
                   2929: 
                   2930: Both! After packaging the nasty details in macro definitions we
                   2931: realized that we could switch between direct and indirect threading by
                   2932: simply setting a compilation flag (@code{-DDIRECT_THREADED}) and
                   2933: defining a few machine-specific macros for the direct-threading case.
                   2934: On the Forth level we also offer access words that hide the
                   2935: differences between the threading methods (@pxref{Threading Words}).
                   2936: 
                   2937: Indirect threading is implemented completely
                   2938: machine-independently. Direct threading needs routines for creating
                   2939: jumps to the executable code (e.g. to docol or dodoes). These routines
                   2940: are inherently machine-dependent, but they do not amount to many source
                   2941: lines. I.e., even porting direct threading to a new machine is a small
                   2942: effort.
                   2943: 
1.4       anton    2944: @node DOES>,  , Direct or Indirect Threaded?, Threading
1.3       anton    2945: @subsection DOES>
                   2946: One of the most complex parts of a Forth engine is @code{dodoes}, i.e.,
                   2947: the chunk of code executed by every word defined by a
                   2948: @code{CREATE}...@code{DOES>} pair. The main problem here is: How to find
                   2949: the Forth code to be executed, i.e. the code after the @code{DOES>} (the
                   2950: DOES-code)? There are two solutions:
                   2951: 
                   2952: In fig-Forth the code field points directly to the dodoes and the
                   2953: DOES-code address is stored in the cell after the code address
                   2954: (i.e. at cfa cell+). It may seem that this solution is illegal in the
                   2955: Forth-79 and all later standards, because in fig-Forth this address
                   2956: lies in the body (which is illegal in these standards). However, by
                   2957: making the code field larger for all words this solution becomes legal
                   2958: again. We use this approach for the indirect threaded version. Leaving
                   2959: a cell unused in most words is a bit wasteful, but on the machines we
                   2960: are targetting this is hardly a problem. The other reason for having a
                   2961: code field size of two cells is to avoid having different image files
1.4       anton    2962: for direct and indirect threaded systems (@pxref{System Architecture}).
1.3       anton    2963: 
                   2964: The other approach is that the code field points or jumps to the cell
                   2965: after @code{DOES}. In this variant there is a jump to @code{dodoes} at
                   2966: this address. @code{dodoes} can then get the DOES-code address by
                   2967: computing the code address, i.e., the address of the jump to dodoes,
                   2968: and add the length of that jump field. A variant of this is to have a
                   2969: call to @code{dodoes} after the @code{DOES>}; then the return address
                   2970: (which can be found in the return register on RISCs) is the DOES-code
                   2971: address. Since the two cells available in the code field are usually
                   2972: used up by the jump to the code address in direct threading, we use
                   2973: this approach for direct threading. We did not want to add another
                   2974: cell to the code field.
                   2975: 
1.4       anton    2976: @node Primitives, System Architecture, Threading, Internals
1.3       anton    2977: @section Primitives
                   2978: 
1.4       anton    2979: @menu
                   2980: * Automatic Generation::        
                   2981: * TOS Optimization::            
                   2982: * Produced code::               
                   2983: @end menu
                   2984: 
                   2985: @node Automatic Generation, TOS Optimization, Primitives, Primitives
1.3       anton    2986: @subsection Automatic Generation
                   2987: 
                   2988: Since the primitives are implemented in a portable language, there is no
                   2989: longer any need to minimize the number of primitives. On the contrary,
                   2990: having many primitives is an advantage: speed. In order to reduce the
                   2991: number of errors in primitives and to make programming them easier, we
                   2992: provide a tool, the primitive generator (@file{prims2x.fs}), that
                   2993: automatically generates most (and sometimes all) of the C code for a
                   2994: primitive from the stack effect notation.  The source for a primitive
                   2995: has the following form:
                   2996: 
                   2997: @format
                   2998: @var{Forth-name}       @var{stack-effect}      @var{category}  [@var{pronounc.}]
                   2999: [@code{""}@var{glossary entry}@code{""}]
                   3000: @var{C code}
                   3001: [@code{:}
                   3002: @var{Forth code}]
                   3003: @end format
                   3004: 
                   3005: The items in brackets are optional. The category and glossary fields
                   3006: are there for generating the documentation, the Forth code is there
                   3007: for manual implementations on machines without GNU C. E.g., the source
                   3008: for the primitive @code{+} is:
                   3009: @example
                   3010: +    n1 n2 -- n    core    plus
                   3011: n = n1+n2;
                   3012: @end example
                   3013: 
                   3014: This looks like a specification, but in fact @code{n = n1+n2} is C
                   3015: code. Our primitive generation tool extracts a lot of information from
                   3016: the stack effect notations@footnote{We use a one-stack notation, even
                   3017: though we have separate data and floating-point stacks; The separate
                   3018: notation can be generated easily from the unified notation.}: The number
                   3019: of items popped from and pushed on the stack, their type, and by what
                   3020: name they are referred to in the C code. It then generates a C code
                   3021: prelude and postlude for each primitive. The final C code for @code{+}
                   3022: looks like this:
                   3023: 
                   3024: @example
                   3025: I_plus:        /* + ( n1 n2 -- n ) */  /* label, stack effect */
                   3026: /*  */                          /* documentation */
1.4       anton    3027: @{
1.3       anton    3028: DEF_CA                          /* definition of variable ca (indirect threading) */
                   3029: Cell n1;                        /* definitions of variables */
                   3030: Cell n2;
                   3031: Cell n;
                   3032: n1 = (Cell) sp[1];              /* input */
                   3033: n2 = (Cell) TOS;
                   3034: sp += 1;                        /* stack adjustment */
                   3035: NAME("+")                       /* debugging output (with -DDEBUG) */
1.4       anton    3036: @{
1.3       anton    3037: n = n1+n2;                      /* C code taken from the source */
1.4       anton    3038: @}
1.3       anton    3039: NEXT_P1;                        /* NEXT part 1 */
                   3040: TOS = (Cell)n;                  /* output */
                   3041: NEXT_P2;                        /* NEXT part 2 */
1.4       anton    3042: @}
1.3       anton    3043: @end example
                   3044: 
                   3045: This looks long and inefficient, but the GNU C compiler optimizes quite
                   3046: well and produces optimal code for @code{+} on, e.g., the R3000 and the
                   3047: HP RISC machines: Defining the @code{n}s does not produce any code, and
                   3048: using them as intermediate storage also adds no cost.
                   3049: 
                   3050: There are also other optimizations, that are not illustrated by this
                   3051: example: Assignments between simple variables are usually for free (copy
                   3052: propagation). If one of the stack items is not used by the primitive
                   3053: (e.g.  in @code{drop}), the compiler eliminates the load from the stack
                   3054: (dead code elimination). On the other hand, there are some things that
                   3055: the compiler does not do, therefore they are performed by
                   3056: @file{prims2x.fs}: The compiler does not optimize code away that stores
                   3057: a stack item to the place where it just came from (e.g., @code{over}).
                   3058: 
                   3059: While programming a primitive is usually easy, there are a few cases
                   3060: where the programmer has to take the actions of the generator into
                   3061: account, most notably @code{?dup}, but also words that do not (always)
                   3062: fall through to NEXT.
                   3063: 
1.4       anton    3064: @node TOS Optimization, Produced code, Automatic Generation, Primitives
1.3       anton    3065: @subsection TOS Optimization
                   3066: 
                   3067: An important optimization for stack machine emulators, e.g., Forth
                   3068: engines, is keeping  one or more of the top stack items in
1.4       anton    3069: registers.  If a word has the stack effect @var{in1}...@var{inx} @code{--}
                   3070: @var{out1}...@var{outy}, keeping the top @var{n} items in registers
1.3       anton    3071: @itemize
                   3072: @item
                   3073: is better than keeping @var{n-1} items, if @var{x>=n} and @var{y>=n},
                   3074: due to fewer loads from and stores to the stack.
                   3075: @item is slower than keeping @var{n-1} items, if @var{x<>y} and @var{x<n} and
                   3076: @var{y<n}, due to additional moves between registers.
                   3077: @end itemize
                   3078: 
                   3079: In particular, keeping one item in a register is never a disadvantage,
                   3080: if there are enough registers. Keeping two items in registers is a
                   3081: disadvantage for frequent words like @code{?branch}, constants,
                   3082: variables, literals and @code{i}. Therefore our generator only produces
                   3083: code that keeps zero or one items in registers. The generated C code
                   3084: covers both cases; the selection between these alternatives is made at
                   3085: C-compile time using the switch @code{-DUSE_TOS}. @code{TOS} in the C
                   3086: code for @code{+} is just a simple variable name in the one-item case,
                   3087: otherwise it is a macro that expands into @code{sp[0]}. Note that the
                   3088: GNU C compiler tries to keep simple variables like @code{TOS} in
                   3089: registers, and it usually succeeds, if there are enough registers.
                   3090: 
                   3091: The primitive generator performs the TOS optimization for the
                   3092: floating-point stack, too (@code{-DUSE_FTOS}). For floating-point
                   3093: operations the benefit of this optimization is even larger:
                   3094: floating-point operations take quite long on most processors, but can be
                   3095: performed in parallel with other operations as long as their results are
                   3096: not used. If the FP-TOS is kept in a register, this works. If
                   3097: it is kept on the stack, i.e., in memory, the store into memory has to
                   3098: wait for the result of the floating-point operation, lengthening the
                   3099: execution time of the primitive considerably.
                   3100: 
                   3101: The TOS optimization makes the automatic generation of primitives a
                   3102: bit more complicated. Just replacing all occurrences of @code{sp[0]} by
                   3103: @code{TOS} is not sufficient. There are some special cases to
                   3104: consider:
                   3105: @itemize
                   3106: @item In the case of @code{dup ( w -- w w )} the generator must not
                   3107: eliminate the store to the original location of the item on the stack,
                   3108: if the TOS optimization is turned on.
1.4       anton    3109: @item Primitives with stack effects of the form @code{--}
                   3110: @var{out1}...@var{outy} must store the TOS to the stack at the start.
                   3111: Likewise, primitives with the stack effect @var{in1}...@var{inx} @code{--}
1.3       anton    3112: must load the TOS from the stack at the end. But for the null stack
                   3113: effect @code{--} no stores or loads should be generated.
                   3114: @end itemize
                   3115: 
1.4       anton    3116: @node Produced code,  , TOS Optimization, Primitives
1.3       anton    3117: @subsection Produced code
                   3118: 
                   3119: To see what assembly code is produced for the primitives on your machine
                   3120: with your compiler and your flag settings, type @code{make engine.s} and
1.4       anton    3121: look at the resulting file @file{engine.s}.
1.3       anton    3122: 
1.17      anton    3123: @node System Architecture, Performance, Primitives, Internals
1.3       anton    3124: @section System Architecture
                   3125: 
                   3126: Our Forth system consists not only of primitives, but also of
                   3127: definitions written in Forth. Since the Forth compiler itself belongs
                   3128: to those definitions, it is not possible to start the system with the
                   3129: primitives and the Forth source alone. Therefore we provide the Forth
                   3130: code as an image file in nearly executable form. At the start of the
                   3131: system a C routine loads the image file into memory, sets up the
                   3132: memory (stacks etc.) according to information in the image file, and
                   3133: starts executing Forth code.
                   3134: 
                   3135: The image file format is a compromise between the goals of making it
                   3136: easy to generate image files and making them portable. The easiest way
                   3137: to generate an image file is to just generate a memory dump. However,
                   3138: this kind of image file cannot be used on a different machine, or on
                   3139: the next version of the engine on the same machine, it even might not
                   3140: work with the same engine compiled by a different version of the C
                   3141: compiler. We would like to have as few versions of the image file as
                   3142: possible, because we do not want to distribute many versions of the
                   3143: same image file, and to make it easy for the users to use their image
                   3144: files on many machines. We currently need to create a different image
                   3145: file for machines with different cell sizes and different byte order
1.17      anton    3146: (little- or big-endian)@footnote{We are considering adding information to the
1.3       anton    3147: image file that enables the loader to change the byte order.}.
                   3148: 
                   3149: Forth code that is going to end up in a portable image file has to
1.4       anton    3150: comply to some restrictions: addresses have to be stored in memory with
                   3151: special words (@code{A!}, @code{A,}, etc.) in order to make the code
                   3152: relocatable. Cells, floats, etc., have to be stored at the natural
                   3153: alignment boundaries@footnote{E.g., store floats (8 bytes) at an address
                   3154: dividable by~8. This happens automatically in our system when you use
                   3155: the ANS Forth alignment words.}, in order to avoid alignment faults on
                   3156: machines with stricter alignment. The image file is produced by a
                   3157: metacompiler (@file{cross.fs}).
1.3       anton    3158: 
                   3159: So, unlike the image file of Mitch Bradleys @code{cforth}, our image
                   3160: file is not directly executable, but has to undergo some manipulations
                   3161: during loading. Address relocation is performed at image load-time, not
                   3162: at run-time. The loader also has to replace tokens standing for
                   3163: primitive calls with the appropriate code-field addresses (or code
                   3164: addresses in the case of direct threading).
1.4       anton    3165: 
1.17      anton    3166: @node  Performance,  , System Architecture, Internals
                   3167: @section Performance
                   3168: 
                   3169: On RISCs the Gforth engine is very close to optimal; i.e., it is usually
                   3170: impossible to write a significantly faster engine.
                   3171: 
                   3172: On register-starved machines like the 386 architecture processors
                   3173: improvements are possible, because @code{gcc} does not utilize the
                   3174: registers as well as a human, even with explicit register declarations;
                   3175: e.g., Bernd Beuster wrote a Forth system fragment in assembly language
                   3176: and hand-tuned it for the 486; this system is 1.19 times faster on the
                   3177: Sieve benchmark on a 486DX2/66 than Gforth compiled with
                   3178: @code{gcc-2.6.3} with @code{-DFORCE_REG}.
                   3179: 
                   3180: However, this potential advantage of assembly language implementations
                   3181: is not necessarily realized in complete Forth systems: We compared
                   3182: Gforth (compiled with @code{gcc-2.6.3} and @code{-DFORCE_REG}) with
1.18    ! anton    3183: Win32Forth 1.2093 and LMI's NT Forth (Beta, May 1994), two systems
        !          3184: written in assembly, and with two systems written in C: PFE-0.9.11
        !          3185: (compiled with @code{gcc-2.6.3} with the default configuration for
        !          3186: Linux: @code{-O2 -fomit-frame-pointer -DUSE_REGS}) and ThisForth Beta
        !          3187: (compiled with gcc-2.6.3 -O3 -fomit-frame-pointer). We benchmarked
        !          3188: Gforth, PFE and ThisForth on a 486DX2/66 under Linux. Kenneth O'Heskin
        !          3189: kindly provided the results for Win32Forth and NT Forth on a 486DX2/66
        !          3190: with similar memory performance under Windows NT.
1.17      anton    3191:  
                   3192: We used four small benchmarks: the ubiquitous Sieve; bubble-sorting and
                   3193: matrix multiplication come from the Stanford integer benchmarks and have
                   3194: been translated into Forth by Martin Fraeman; we used the versions
                   3195: included in the TILE Forth package; and a recursive Fibonacci number
                   3196: computation for benchmark calling performance. The following table shows
                   3197: the time taken for the benchmarks scaled by the time taken by Gforth (in
                   3198: other words, it shows the speedup factor that Gforth achieved over the
                   3199: other systems).
                   3200: 
                   3201: @example
                   3202: relative             Win32-        NT               This-
                   3203:   time     Gforth     Forth     Forth       PFE     Forth
                   3204: sieve        1.00      1.30      1.07      1.67      2.98
                   3205: bubble       1.00      1.30      1.40      1.66
                   3206: matmul       1.00      1.40      1.29      2.24
                   3207: fib          1.00      1.44      1.26      1.82      2.82
                   3208: @end example
                   3209: 
                   3210: You may find the good performance of Gforth compared with the systems
                   3211: written in assembly language quite surprising. One important reason for
                   3212: the disappointing performance of these systems is probably that they are
                   3213: not written optimally for the 486 (e.g., they use the @code{lods}
                   3214: instruction). In addition, Win32Forth uses a comfortable, but costly
                   3215: method for relocating the Forth image: like @code{cforth}, it computes
                   3216: the actual addresses at run time, resulting in two address computations
                   3217: per NEXT (@pxref{System Architecture}).
                   3218: 
                   3219: The speedup of Gforth over PFE and ThisForth can be easily explained
                   3220: with the self-imposed restriction to standard C (although the measured
                   3221: implementation of PFE uses a GNU C extension: global register
                   3222: variables), which makes efficient threading impossible.  Moreover,
                   3223: current C compilers have a hard time optimizing other aspects of the
                   3224: ThisForth source.
                   3225: 
                   3226: Note that the performance of Gforth on 386 architecture processors
                   3227: varies widely with the version of @code{gcc} used. E.g., @code{gcc-2.5.8}
                   3228: failed to allocate any of the virtual machine registers into real
                   3229: machine registers by itself and would not work correctly with explicit
                   3230: register declarations, giving a 1.3 times slower engine (on a 486DX2/66
                   3231: running the Sieve) than the one measured above.
                   3232: 
1.4       anton    3233: @node Bugs, Pedigree, Internals, Top
                   3234: @chapter Bugs
                   3235: 
1.17      anton    3236: Known bugs are described in the file BUGS in the Gforth distribution.
                   3237: 
                   3238: If you find a bug, please send a bug report to !!. A bug report should
                   3239: describe the Gforth version used (it is announced at the start of an
                   3240: interactive Gforth session), the machine and operating system (on Unix
                   3241: systems you can use @code{uname -a} to produce this information), the
                   3242: installation options (!! a way to find them out), and a complete list of
                   3243: changes you (or your installer) have made to the Gforth sources (if
                   3244: any); it should contain a program (or a sequence of keyboard commands)
                   3245: that reproduces the bug and a description of what you think constitutes
                   3246: the buggy behaviour.
                   3247: 
                   3248: For a thorough guide on reporting bugs read @ref{Bug Reporting, , How
                   3249: to Report Bugs, gcc.info, GNU C Manual}.
                   3250: 
                   3251: 
1.4       anton    3252: @node Pedigree, Word Index, Bugs, Top
                   3253: @chapter Pedigree
                   3254: 
1.17      anton    3255: Gforth descends from BigForth (1993) and fig-Forth. Gforth and PFE (by
                   3256: Dirk Zoller) will cross-fertilize each other. Of course, a significant part of the design of Gforth was prescribed by ANS Forth.
                   3257: 
                   3258: Bernd Paysan wrote BigForth, a child of VolksForth.
                   3259: 
                   3260: VolksForth descends from F83. !! Authors? When?
                   3261: 
                   3262: Laxen and Perry wrote F83 as a model implementation of the
                   3263: Forth-83 standard. !! Pedigree? When?
                   3264: 
                   3265: A team led by Bill Ragsdale implemented fig-Forth on many processors in
                   3266: 1979. Dean Sanderson and Bill Ragsdale developed the original
                   3267: implementation of fig-Forth based on microForth.
                   3268: 
                   3269: !! microForth pedigree
                   3270: 
                   3271: A part of the information in this section comes from @cite{The Evolution
                   3272: of Forth} by Elizabeth D. Rather, Donald R. Colburn and Charles
                   3273: H. Moore, presented at the HOPL-II conference and preprinted in SIGPLAN
                   3274: Notices 28(3), 1993.  You can find more historical and genealogical
                   3275: information about Forth there.
                   3276: 
1.4       anton    3277: @node Word Index, Node Index, Pedigree, Top
                   3278: @chapter Word Index
                   3279: 
1.18    ! anton    3280: This index is as incomplete as the manual. Each word is listed with
        !          3281: stack effect and wordset.
1.17      anton    3282: 
                   3283: @printindex fn
                   3284: 
1.4       anton    3285: @node Node Index,  , Word Index, Top
                   3286: @chapter Node Index
1.17      anton    3287: 
                   3288: This index is even less complete than the manual.
1.1       anton    3289: 
                   3290: @contents
                   3291: @bye
                   3292: 

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>